WO2023277863A1

WO2023277863A1 - Print agent coverage amounts in additive manufacturing

Info

Publication number: WO2023277863A1
Application number: PCT/US2021/039359
Authority: WO
Inventors: Arnau CODINA SABORIT; Jordi BAUTISTA BALLESTER; Albert RIPOLL OLIVERAS
Original assignee: Hewlett-Packard Development Company, L.P.
Priority date: 2021-06-28
Filing date: 2021-06-28
Publication date: 2023-01-05

Abstract

A method includes generating a plurality of calibration objects using additive manufacturing in a plurality of layers in which other objects are generated, wherein the other objects fill a predetermined proportion of the layers, and the objects are generated using a predetermined coverage amount of a cooling agent. A measurement of a physical property of each calibration object may be obtained by processing circuitry. The measurements may include a first set of measurements of calibration objects generated with different filled proportions of the layers, and a second set of measurements of calibration objects generated with different coverage amounts. The method may include characterising, by processing circuitry and based on the sets of measurements, a relationship between cooling agent coverage amounts and an indication of an amount of radiation received by the calibration object which is reflected from build material during object generation, the relationship useful in generating subsequent objects.

Description

PRINT AGENT COVERAGE AMOUNTS IN ADDITIVE MANUFACTURING

BACKGROUND

[0001] Additive manufacturing techniques may generate a three-dimensional object through the solidification of a build material, for example on a layer-by-layer basis. In examples of such techniques, build material may be supplied in a layer-wise manner and the solidification method may include heating the layers of build material to cause melting in selected regions. In other techniques, chemical solidification and/or binding methods may be used.

BRIEF DESCRIPTION OF THE DRAWINGS

[0002] Non-limiting examples will now be described with reference to the accompanying drawings, in which:

[0003] Figure 1 is an example method of characterising a relationship between cooling agent coverage amounts and reflected radiation for use in additive manufacturing;

[0004] Figures 2A and 2B show examples of arrangements of objects on a print bed including calibration objects;

[0005] Figure 3 is an example showing how object weight may be related to packing density;

[0006] Figure 4 is an example of a method for use in characterising a relationship for use in additive manufacturing;

[0007] Figures 5A and 5B show examples demonstrating how reflected energy may impact calibration objects;

[0008] Figure 6 shows an example of a conversion which may be applied to cooling agent amounts;

[0009] Figure 7 is an example method of determining cooling agent amounts;

[0010] Figure 8 is an example of an apparatus;

[0011] Figure 9 is another example of an apparatus; and [0012] Figure 10 is an example machine-readable medium associated with a processor.

DETAILED DESCRIPTION

[0013] Additive manufacturing techniques may generate a three-dimensional object through the solidification of a build material. In some examples, the build material is a powder-like granular material, which may for example be a plastic, ceramic or metal powder and the properties of generated objects may depend on the type of build material and the type of solidification mechanism used. Build material may be deposited, for example on a print bed and processed layer by layer, for example within a fabrication chamber. According to one example, a suitable build material may be PA12 build material commercially referred to as V1 R10A “HP PA12” available from HP Inc. Other example build materials comprise PA11 material, commercially referred to as HP PA11 available from HP Inc, Thermoplastic Polyurethane (TPU) materials, Thermoplastic Polyamide materials (TPA), and the like.

[0014] In some examples, selective solidification is achieved through directional application of energy, for example using a laser or electron beam which results in solidification of build material where the directional energy is applied. In some examples, at least one print agent may be selectively applied to the build material, and may be liquid when applied. For example, a fusing agent (also termed a ‘coalescence agent’ or ‘coalescing agent’) may be selectively distributed onto portions of a layer of build material in a pattern derived from data representing a slice of a three-dimensional object to be generated (which may for example be determined from structural design data). The data may be derived from a digital or data model of the object. The fusing agent may have a composition which absorbs energy such that, when energy (for example, heat) is applied to the layer, the build material to which it has been applied heats up, coalesces and solidifies, upon cooling, to form a slice of the three-dimensional object in accordance with the pattern. In other examples, coalescence may be achieved in some other manner.

[0015] According to one example, a suitable fusing agent may be an ink-type formulation comprising carbon black, such as, for example, the fusing agent formulation commercially referred to as V1Q60A “HP fusing agent” available from HP Inc. Such a fusing agent may comprise any or any combination of an infra-red light absorber, a near infra-red light absorber, a visible light absorber and a UV light absorber. Examples of fusing agents comprising visible light absorption enhancers are dye based colored ink and pigment based colored ink, such as inks commercially referred to as CE039A and CE042A available from HP Inc.

[0016] In addition to a fusing agent, in some examples, a print agent may comprise a coalescence modifier agent, which acts to modify the effects of a fusing agent for example by reducing or increasing coalescence or to assist in producing a particular finish or appearance to an object, and such agents may therefore be termed detailing agents. In some examples, a detailing agent may comprise a cooling agent and may be used near edge surfaces of an object being printed to reduce coalescence, for example extending in a perimeter about the surface of an object in a layer. According to one example, a suitable cooling agent may be a formulation commercially referred to as V1Q61A “HP detailing agent” available from HP Inc.

[0017] A coloring agent, for example comprising a dye or colorant, may in some examples be used as a fusing agent or a coalescence modifier agent, and/or as a print agent to provide a particular color for the object.

[0018] As noted above, additive manufacturing systems may generate objects based on structural design data. This may involve a designer determining a data model of an object to be generated, for example using a computer aided design (CAD) application. The model may define the solid portions of the object. To generate a three-dimensional object from the model using an additive manufacturing system, the model data can be processed to define slices or parallel planes of the model. Each slice may define a portion of a respective layer of build material that is to be solidified or caused to coalesce by the additive manufacturing system.

[0019] During object generation, at least one energy source, for example infrared source(s), heat lamp(s), array(s) of LEDs, and the like, which may be static and/or scanned over the surface of a layer of build material on a print bed, may be used to provide energy.

[0020] It has been noted that the temperature of a layer of build material being processed has a relationship with the amount of build material on which no fusing agent has been applied. While temperatures may be expected to increase when there is more fusing agent applied, it has been noted that temperatures could also increase with the amount of build material which is not treated with fusing agent. Without wishing to be bound by theory, this may be because, while fusing agent tends to absorb incident energy, build material which is not treated with fusing agent tends to be at least partially reflective. As the energy sources are often equipped with reflectors and/or concentrators, reflected energy is ‘re-reflected’ back towards the print bed. This ‘re-reflected’ or ‘re-radiated’ energy may therefore be higher in the case of a layer with a high proportion of build material which is not treated with fusing agent than in the case of a layer with a low proportion of build material which Is not treated with fusing agent.

[0021] It has been noted that, as the solidified proportion (the ‘packing density’) of a layer increases, and the re-radiation effect decreases, objects tend to decrease in weight indicating a lower degree of fusing. In some examples, this re-radiation effect is somewhat localised, i.e. the radiated energy may be re-reflected towards the general area of the print bed from which it was reflected (allowing for some dispersion and imperfections in reflective surfaces), although the size of the affected region may depend on the configuration of the apparatus in use.

[0022] Figure 1 is an example of a method of characterising a relationship between cooling agent coverage amounts and re-radiation for object portions in a layer. In this example the method is carried out in part by processing circuitry, which may comprise at least one processor.

[0023] The method comprises, in block 102, generating a plurality of calibration objects using additive manufacturing. In this example, the objects are generated in a plurality of layers in which other objects are generated, wherein the other objects fill a predetermined proportion of the layers, and wherein the calibration objects and the other objects are generated using a predetermined coverage amount of a cooling agent.

[0024] In this example, generating the objects may comprise generating objects in a layer-wise manner by selectively solidifying portions of layers of build material formed on a print bed of an additive manufacturing apparatus. The selective solidification in this example is achieved by selectively applying print agents, for example through use of ‘inkjet’ liquid distribution technologies, and applying energy, for example heat, to each layer using one or more energy source. The placement of such agents may be determined using control instructions, which may be derived from object model data modelling objects to be generated (including the calibration objects.) The object model data may for example comprise a Computer Aided Design (CAD) model, and/or may for example comprise a STereoLithographic (STL) data file. [0025] In some examples, block 102 may comprise generating a plurality of different calibration ‘batches’, i.e., a batch of objects comprising at least one calibration object and at least one other object. Each batch has an associated filled proportion (or packing density) of material to be solidified and an associated coverage amount of cooling agent.

[0026] In examples, the cooling agent may be applied to a region surrounding each object, for example extending from a normal of the surface by a predetermined distance. In other examples the cooling agent may be applied outside the object portion in the layer and within a region corresponding to a ‘boundary box’ enclosing the slice of the object corresponding to that layer, and in some examples being larger than the object slice, and which may for example be rectangular in shape. In some examples, the boundary box may be defined to enclose an object model in three dimensions, and each slice may comprise a slice of at least one such boundary box. In other examples, the cooling agent may be applied to any portion of the layer which is not to be fused.

[0027] In a given calibration batch, a calibration object may be generated over a number of layers. Over those same layers, at least one other object, which may be termed a ‘filler object' herein, is also generated.

[0028] While such filler objects may in principle comprise any shape or form, in some examples they individually occupy a relatively small proportion of a layer and are straight-sided, such that they have a consistent cross-sectional profile over the number of layers. For example, a filler object may comprise a cuboid having a cross sectional area which is less than 5%, or less than 3%, or less than 1% of the cross sectional area of a print bed of the additive manufacturing apparatus. Providing straight sided filler objects means that the proportion of the print bed filled by the filler objects remains consistent over the number of layers in which the calibration object is generated. In addition, providing relatively small filler objects means that the filler objects may be well dispersed over the surface area. However, in principle, a single filler object may be generated to provide the filled proportion of the layers of the calibration batch.

[0029] In some examples, the calibration object may comprise an object having a lattice body. In other words, the object body is intended to comprise struts formed around a plurality of voids. This serves to reduce the amount of build material which forms part of the object (noting that, in some examples, unfused build material may be recycled in subsequent additive manufacturing operations). In addition, such an object may be relatively variable depending on the extent of fusion. For example, if sufficient energy is absorbed to cause ‘over fusing’, the voids may at least partially close up. However, if insufficient energy is absorbed, at least some of the struts may fail to form. In some examples, the structure may comprise a ‘sandwich’ arrangement of different lattice densities. For example, the first few layers with which the calibration object is formed may comprise a relatively dense lattice structure, comprising a relatively high number of struts and/or relatively thick struts. This may be followed by a number of additional layers in which the lattice structure is relatively less dense, comprising fewer and/or narrower struts, and then a final few layers having the dense lattice structure of the first few layers. Providing these relatively dense layers may ensure that the calibration object is relatively robust while still having significant voids in the intervening layers.

[0030] In some examples, the calibration object may comprise an identifier, for example formed embossed therein or marked on the surface using colorants or the like.

[0031] The calibration batches may be generated at the same height within an additive manufacturing operation, at different heights within the same additive manufacturing operation, and/or in different additive manufacturing operations. For example, an additive manufacturing apparatus may be used to generate a first calibration batch, generated objects and any unused build material may be removed therefrom, and the additive manufacturing apparatus may be reused to generate the objects of a second calibration batch. In some examples, more than one calibration batch may be printed in the same set of layers. For example, one half of the print bed may be treated with a first amount of cooling agent and another half of the print bed may be treated with a second amount of cooling agent, each half comprising a calibration object such that two calibration batches are generated in the same set of layers of additive manufacturing.

[0032] Block 104 of Figure 1 comprises obtaining, by processing circuitry, a measurement of a physical property of each calibration object. The measurements include a first set of measurements of calibration objects generated with different fill proportions of the layers, and a second set of measurements of calibration objects generated with different coverage amounts. Measurements may belong to one or both sets. For example, a first calibration batch may be generated with a first coverage amount of cooling agent and a first filled proportion and a second calibration batch may be generated with the second coverage amount of cooling agent and a second filled proportion. [0033] For example, block 104 may comprise directly measuring the physical property, or receiving an indication of a measurement. The measurement may for example comprise a measurement of a physical property which provides a proxy for a degree of fusing undergone, for example, a weight or mass. Generally, in examples herein, a higher extent of fusion (which may in turn be associated with a higher level of energy absorption, which may be due, at least in part, to the effect of re-radiation of energy) is associated with a greater weight.

[0034] While weights are used in some examples herein, other measurements may be used as a proxy for a degree of fusion undergone. For example, the measurements may comprise a strength or flexibility of the object(s) (a higher extent of fusion may be associated with a higher resistance to breaking in a strength test, and/or a lower amount of flexibility) and/or a density of the object (which may in some examples be inferred from the weight). In some cases, an object which is subjected to a high extent of fusing may ‘grow’ as build material which is at least partially fused may be incorporated therein, or unfused build material may adhere to the surfaces thereof. Therefore, an increase in dimensions may be associated with an increased degree of fusion and the measurements may comprise measurements of dimensions of the calibration objects. However, in other examples, a high extent of fusion may be associated with a higher degree of shrinkage on cooling and thus a reduction in dimensions may be associated with a high extent of fusion. This can depend on factors such as, for example, the materials used.

[0035] In still further examples, small features may be present or absent depending on the extent of fusion undergone by the object. For example, small holes may close up when the extent of fusion of the object is high whereas the holes may remain open when the extent of fusion is lower. For example, holes of varying sizes could be specified in object model data and used to indicate an extent of fusion, for example based on the smallest hole that remains open. Protrusions which are specified in object model data (which may be the basis on which the object is generated) may be absent at lower extents or degrees of fusion but present at higher degrees of fusion. For example, the smallest protruding feature which is successfully generated in a series of features of increasing size may give an indication of the extent of fusion, and the measurement may comprise a measurement to identify this feature. Other examples of measurable properties of generated objects may also provide proxies for the extent of fusion of the build material and/or may provide the measurements of block 104. [0036] Block 106 comprises characterising, by processing circuitry (which may be the same processing circuitry as was used to carry out block 104) and based on the first and second set of measurements, a relationship between cooling agent coverage amounts and an indication of an amount of radiation received by calibration objects that was (i.e. was or is estimated to be) reflected from build material during object generation (i.e. the amount of ‘re-radiated’ energy), wherein the relationship is for use in generating subsequent objects, for example an object having an intended physical property, which may be the same as the property measured to obtain the measurements of block 104.

[0037] As noted above, an increased amount of re-radiation may be associated with a relatively low packing density, and in particular with a low local packing density in the case that the re-radiation effect is relatively localised, as described above. Moreover, an increased amount of re-radiation may be associated with an increased extent of fusion. However, this may be at least in part countered by using an increased coverage amount of cooling agent. Therefore, by characterising a relationship between cooling agent coverage amounts and an indication of the extent to which an object is affected by re-radiation, an appropriate amount of cooling agent may be used in subsequent build operations in order to counter an increased fusion associated with re-radiation in the case of a lower packing density (or, in examples, a lower local packing density), and/or to counter a lower degree of fusion associated with a higher packing density (or, in examples, a higher local packing density).

[0038] Figure 2A shows an example of a first arrangement of objects on a print bed 200. In this example, two calibration batches are included in the arrangement. The left-hand 202 of a print bed 200 is associated with a first cooling agent coverage amount and the right-hand 204 of the print bed 200 is associated with a second cooling agent amount. Each calibration batch comprises a calibration object 206a, 206b, which in this example comprise lattice objects, and a number of filler objects 208 (only some of which are labelled), in this example cuboids which have nominally the same dimensions as one another (i.e. are generated based on the same object model data), and are formed over the same number of layers as the calibration objects 206.

[0039] The calibration objects 206 generally have the same size and shape as each other, and comprise lattice objects having a tablet-, or slab-like form, i.e. having greater X and Y dimensions in the print bed than a Z, or height dimension, perpendicular to the plane of the print bed. The calibration objects 206 may be generally similar to one another and in some examples may be nominally the same (i.e. generated based on the same object model data), or nominally the same with the exception of an identifier, which may for example be formed embossed therein or marked on the surface using colourants or the like.

[0040] In this example, the filler objects 208 are randomly distributed to provide a predetermined filled proportion, or packing density. For example, a packing density may be around 20%, i.e., 20% of the surface area of the calibration batch is to be solidified whereas 80% is left unsolidified.

[0041] The left-hand side 202 of the print bed 200 is associated with a cooling agent coverage amount at X% of a nominal amount whereas the right-hand side of the print bed is associated with a cooling agent coverage amount of Y% of the nominal amount. X and Y may be different, for example being selected from a range of 50% to 200% of the nominal amount. In some examples, the range may be set considering the materials in use. For example, with some build materials, if less than a threshold amount of cooling agent is used, the build material may tend to ‘cake’ when heated, even when it is not intended for that portion of build material to fuse. This may reduce the recyclability of the unfused build material, and/or unfused build material may adhere to objects being generated, causing defects or adding to cleaning tasks. Therefore, the lower end of the range may be selected in order to avoid such caking.

[0042] In this example, cooling agent is printed around a perimeter of each surface of an object, extending a predetermined distance in a direction that is normal to the surface. Therefore, some cooling agent may be printed in the voids inside the lattice calibration object 206, around the outer boundary thereof, and around the outer boundary of the test objects 208. Where lattice structs are separated by less the predetermined distance, cooling agent may be applied to fill the voids between struts.

[0043] Figure 2B shows an example of a second arrangement of objects on a print bed 200. As in Figure 2A, two calibration batches are included in the arrangement. The left-hand 202 of the print bed 200 is associated with a first cooling agent coverage amount (P% of the nominal amount) and the right-hand 204 of the print bed 200 is associated with a second cooling agent amount (Q% of the nominal amount). P and Q may be different from one another, and may be in the range of 50 to 200% of the nominal amount. In some examples, the value of P is equal to the value of X and the value of Q is equal to the value of Y. In some examples, the left-hand side of the print bed may always be printed with a given proportion of the nominal amount of cooling agent and the right-hand side of the print bed may always be printed with a different given proportion of the nominal amount of cooling agent.

[0044] The second arrangement of objects is generally similar to the first arrangement of objects but the packing density is lower - there are 16 filler objects arranged on each side of the print bed, compared to 20 filler objects 208 on each side in the first arrangement. The filler objects 208 are again arranged randomly across the surface of the print bed but, in this example, the calibration objects 206c, 206d occupy the same position within the print bed as in the first arrangement. This means that, but for the difference in fill factor and associated re-radiation and a possible difference in cooling agent coverage amount, the object generation parameters for all four calibration objects 206a-d are broadly similar.

[0045] In this example, the first and second arrangements are printed at different levels in the same fabrication chamber in a single additive manufacturing operation. In other words, the first arrangement may be printed first and the second arrangement may be printed second. There may be more such arrangements. For example, around 10 different arrangements, each comprising two calibration batches, may be generated in a single additive manufacturing operation. In some examples, each batch may be separated by a number of untreated layers of build material, i.e. build material to which no fusing agent is applied.

[0046] After having been generated, the objects may be removed from the fabrication chamber (for example, being ‘decaked’ from surrounding unfused build material), may in some examples be cleaned, and a physical property of the calibration objects 206 may be measured, which may be a property which provides a proxy for a degree of fusion undergone. In this example, the weights of the calibration objects 206 are measured. Where the calibration objects 206 are provided with identifiers, this may facilitate the method of determining the weights although the identities of the objects could be tracked in some other way. In some examples, any identifier may be selected such that it does not cause a significant change to the weight of one object compared to another.

[0047] In this way, a set of measurements may be determined, i.e. the measurements mentioned in block 104 of Figure 1.

[0048] Each arrangement of objects is associated with a packing density (or filled proportion) and provides two measurements, one for each batch: a first object weight which is associated with that packing density and with a first cooling agent coverage amount and a second object weight which is associated with the packing density and a second cooling agent coverage amount.

[0049] Figure 3 shows the result of measuring the weight of various objects generated in this way. In this example, there is shown a first line associated with circular marks, which is a linear best fit for the weights of calibration objects printed on the left-hand side of a print bed with a cooling agent coverage amount at a contone level of 37.5. In this example, this is a contone level from a scale of 0 to 255, so 37.5 represents just under 15% of the maximum coverage deliverable by this particular apparatus. A second line, associated with square markers, is a linear best fit for the weights of calibration objects printed on the right-hand side of the print bed with a cooling agent coverage amount at a contone level of 75.

[0050] As can be seen, in this case, 20 calibration batches were printed, 10 at each cooling agent coverage level. In this example, there were 10 arrangements of objects similar to those shown in Figures 2A and 2B, the arrangements being associated with different packing densities. One arrangement of two calibration batches was generated at a packing density of each of 3%, 15%, 25% and 30% and two arrangements of two calibration batches were generated at a packing density of each of 5%, 10% and 20%.

[0051] In this example, the specification of the packing density includes the calibration object. However, as the calibration object has a relatively dense top and bottom (a ‘sandwich’ configuration as described above), an average packing density is used over the depth of the object. In other words, taking as an example a batch printed at 10% packing density, each layer to be generated is considered as an array of print addressable pixels, wherein, on average, 10% of the pixels in a layer are to be solidified over the layers of the batch as a whole. While the number of pixels of the cuboid packing objects remains consistent in each layer, the number of pixels which relate to a part of the calibration object may change from layer to layer.

[0052] As can be seen, as packing density increases, the weight tends to decrease, as the reradiation effect also decreases. However, the lower line, which is associated with a greater amount of cooling agent, is generally associated with lower weight objects.

[0053] Figure 3 therefore provides information about how the change in detailing agent impacts object weight. For example, when printing with detailing agent at a contone level of 75 (the line marked with square markers), it may be the case that the batch printed at 10% packing density is predicted by the trend line to result in a calibration object having a nominal, or expected, weight, which is around 7.2 grams. The batch printed with a 15% packing density results in an object which is predicted by the trendline to be ‘underweight’ relative to that nominal weight - in particular, a drop from 7.2g to 7g, or a change in weight of about 2.8%. However, if the DA amount is halved to a contone level of 37.5, this is predicted to restore the object to around its nominal weight.

[0054] In some examples, it may be assumed that the relationship between print agent amount and the percentage change in weight is linear. For example, for every 10 units change in cooling agent contone amount, there may be around 0.8% change in weight. In another example, a percentage change in weight may be associated with a percentage change in cooling agent amount.

[0055] While Figure 3 shows the results for two contone levels, further calibration batches may be generated at other contone level(s) to further characterise the relationship.

[0056] Figure 4 sets out a method of carrying out block 106 of Figure 1 , and may be implemented by processing circuitry. In this example, the method is carried out for each layer of the intended fabrication chamber content.

[0057] In examples, the fabrication chamber content is modelled as a ‘virtual fabrication chamber’. This includes object models (i.e. data models or ‘virtual objects’) for the calibration objects and the filler objects, including their intended position within the fabrication chamber. This virtual fabrication chamber is sliced, each slice corresponding to a layer to be generated in additive manufacturing. Each slice may in effect be a binary bit map or plane, wherein pixels or voxels which are associated with a region within the fabrication chamber which is intended to solidify are associated with one value (for example, 255) and pixels or voxels which are associated with a region of the fabrication chamber which is not intended to solidify are associated with another value (for example, 0).

[0058] In block 402, a radiation kernel is applied to a bitmap representing a slice of the ‘virtual’ build volume modelling the arrangement of objects to be generated to generate a convolved bitmap. While the map of the slice may be a binary bitmap, the convolved bitmap may comprise greyscale values, for example within a range. In one example, the range may be from 0 to 255. For example, in the case that, in the original bit map, pixels to be solidified were associated with a value of 255, the pixels at the surfaces of the object may have a reduced value following convolution, whereas the pixels just outside the object and associated with a value of 0 before convolution may have an increased value following convolution. However, the value of pixels modelling the centre of a relatively large object may remain unchanged if the kernel is smaller than the object cross section.

[0059] The radiation kernel, which may be a gaussian kernel, in effect simulates energy re-reflection on to the layer. Such kernels act as spatial filters which ‘blur’ images, and may in effect operate as a moving window which is scanned across the bitmap. The characteristics of the kernel may be determined or derived for example experimentally, empirically or through application of theory to result in a kernel which produces an intended result based on characteristics of the material(s) being used (e.g. its reflectivity), and the source of the re-radiation (for example, the shape and material of a reflector of a heat source). The radiation kernel may be convolved with the bitmap representing the slice.

[0060] Portions of the layer which are not to be solidified may be associated with a higher amount of reflection of energy, and portions of a layer which are to be solidified may be associated with a lower amount of reflected energy. The blurring provided by the kernel models the tendency for the reflected energy to be somewhat diffused. However, the reflection may be at least somewhat localised- i.e., energy is reflected back to a similar region of the print bed as it was reflected from.

[0061] To discuss this in a little more detail, in an example apparatus, the energy source comprises a strip extending in (nominally) the x direction, which is scanned over the surface of the print bed in (nominally) the y direction. The energy source has a reflector positioned behind it, and the reflector generally reflects energy back towards the local region of the build bed from which it originated. Imperfections in the surface of the reflector, and diffusion, mean that the energy is not perfectly ‘retroreflected’, but is instead somewhat dispersed. Therefore, the kernel in such an example may model the ‘spread’ in reflected energy to an area which extends a few pixels around the point of reflection (where the pixels are the pixels of the bitmap, and may be at a print resolution of the additive manufacturing apparatus). However, in another example apparatus, the spread may be more diffuse, less diffuse, may be asymmetric or the like, and/or may vary in shape over the print bed, according to features such as the shape, material, surface treatment of the reflector, and so one. In some examples, there may be additional components within the apparatus which may cause ‘re-reflection’. The kernel may be designed or selected to reflect such factors. [0062] In block 404, pixels within the convolved map which are associated with the calibration objects are identified. For example, the bitmap of the slice may be used to identify the portions of the convolved map which are representative of the calibration objects. Referring to the examples shown in Figures 2A and 2B, there may be two (or in other examples more) calibration objects identified in this way.

[0063] In block 406, the mean of the pixel values for the portion of the convolved bitmap which is associated with the calibration objects is determined. While there are two objects generated, as these are generated with the same packing density, the mean values for each will be similar. In other words, even though the amount of cooling agent applied to each of the two calibration objects generated in a particular arrangement is different, the amount of re-radiation that each object experiences is similar.

[0064] The mean value may be referred to as a convolution value and is indicative of the extent to which re-radiation is associated with a particular calibration object. This method may be repeated for each slice of the virtual fabrication chamber. In other examples, rather than being a mean value, the convolution value may be another value representative of the effect of the re-radiation on the calibration object, such as a median or the like. In other examples, the amount of re-radiated energy may be determined in some other way, for example by thermal measurements during object generation.

[0065] Each mean value is associated with a packing density, and this allows a relationship between packing density and the convolution value to be characterised.

[0066] Figure 5A and B demonstrate an example of such a relationship, for data corresponding to that shown in Figure 3.

[0067] Figure 5A shows the mean convolution values output by the method of Figure 4 for each slice of the virtual fabrication chamber (which corresponds to an intended layer in additive manufacturing).

[0068] In this example, around 400 layers are processed in generating 20 calibration batches in 10 arrangements of objects. In this example, which is based on the same data as used to generate the graph of Figure 3, the lowermost arrangement of objects has a packing density of 5%, followed by an arrangement at 20%, then 10%, 30%, 5%, 15%, 3%, 25%, 10% and 20% in that order. The packing densities are average packing densities over the layers of generating the objects in each calibration batch. [0069] By way of example, data relating to two arrangements of objects, each generated over around 25 layers, and each comprising two calibration batches with one calibration object in each batch, is highlighted in the graph of Figure 5A. First, the mean value associated with the calibration objects to be generated between layers 10 and 35 (and therefore modelled by slices 10 to 35) is highlighted between dotted lines. This is associated with a packing density of 5%. Second, the mean value associated with the calibration object to be generated between layers 120 and 165 is highlighted between dotted lines. This is associated with a packing density of 30%.

[0070] Where the packing density is higher, the calibration object receives a lower amount of re-radiated energy. In this example, this is associated with a mean value which is generally higher. However, it will be appreciated that this is a unitless value which depends on how the locations of the objects are indicated as values in the original bit map modelling the layer, and this could be reversed in other examples.

[0071] In this example, the calibration objects comprise lattices having the ‘sandwich’ configuration described above. That is, the first and last few layers are formed of a relatively dense lattice compared to a middle section. This results in peaks in the mean convolution value for each calibration object, as the packing density of such layers individually including the material of the lattice is higher than in the middle layers. The value varies from slice to slice according to the actual packing density including the density of the lattice in that slice. If the calibration object was, for example, a solid object, the fluctuations may be relatively small.

[0072] Figure 5B shows the average result over all the layers in which the calibration objects were generated. In this example, the peaks of Figure 5A were excluded, and the average values for the middle portion of the calibration object (i.e. the central layers of the ‘sandwich’) were used to generate the data. This demonstrates a linear relationship between the packing density and the mean values.

[0073] The relationships shown in Figure 2 and Figure 5B can be used to determine a cooling agent compensation to compensate for an effect of re-radiation. In particular, an amount by which the cooling agent coverage should be increased to compensate for the effect of re-radiation may be calculated, as (i) the amount of cooling agent to increase a percentage weight of the calibration object is known and (ii) the relationship between the increase in weight and the convolution value is known. [0074] In order to determine a cooling agent amount to apply in a subsequent build operation, a conversion relative to a nominal amount of cooling agent and a nominal convolution value may be determined. An example conversion relationship is shown in Figure 6.

[0075] A conversion relationship, such as the relationship shown in Figure 6 may be determined on the following basis. In summary, Figure 3 shows (i) a relationship between packing density and weight and (ii) a relationship between cooling agent coverage amounts and weight. Figure 5B shows a relationship between packing density and a calculated value indicative of the re-radiation affecting a particular object. These can be combined to determine a relationship between the amount by which the cooling agent should be adjusted to compensate for a change in weight which would otherwise occur due to re-radiation. It may be noted that the relationships may vary based on, for example, apparatus, apparatus features or operational parameters, print material choices and the like.

[0076] In other examples, a conversion relationship may be determined which is associated with an absolute change in cooling agent coverage amounts, for example a count in contone levels, rather than a scaling value.

[0077] In an example of use of the conversion relationship, it may be assumed that a given print apparatus has been calibrated, or is otherwise determined to produce objects which have intended physical properties for a given packing density using a baseline amount of cooling agent. In a particular example, the nominal packing density may be around 10%.

[0078] Generally speaking, in the conversion relationship set out in Figure 6, for a lower convolution value, which is associated with a lower local packing density, the re-radiation increases and the amount of cooling agent increases to compensate and limit associated ‘over fusing’. For a higher convolution value, which is associated with a higher local packing density, the re-radiation reduces and the amount of cooling agent reduces to compensate. In this example, the reduction in cooling agent is limited to two thirds the nominal amount, as a reduction beyond this point has been seen to be associated with other quality defects for this particular build material such as caking as described above, but in other examples such a threshold need not apply, or there may be different threshold(s).

[0079] To consider a worked example, a printer may be calibrated using a standard set of objects generated at a given packing density, for example 10%. However, when printing a print job of arbitrary content, there will be a different re radiation impact on the objects.

[0080] In an example print job, the packing density in a given layer may be, for example, around 20%. It may therefore be predicted that at least some objects may be generated to have a lower than intended weight, as the effect of the re radiation of energy is reduced.

[0081] In general, a baseline cooling agent coverage amount may be set for a layer of a print job, or the print job as a whole. This baseline may be set following a thermal analysis of the print job or layer, for example being intended to result in an average temperature for the layer or the print job. In some examples, more cooling agent may be specified as the baseline amount when a greater proportion of a fabrication chamber (or a layer thereof) is to be solidified to counter the higher temperatures associated with a greater amount of fusing agent. In other examples, a baseline may be a standard baseline for the print apparatus as a whole, or determined in some other way.

[0082] Figure 7 is an example of a method which uses the determined relationship, and which may be applied to determine an amount of cooling agent to apply in a print job.

[0083] In block 702, the convolution kernel modelling re-radiation described in relation to Figure 4 is applied to a bitmap representing a layer of the print job. In examples herein, the bitmap is a slice of a ‘virtual fabrication chamber’, modelling the intended content of an object generation operation in which at least one object is to be generated. The object or objects may have any specified form (i.e. they may different to the calibration objects and the filler objects used to characterise the relationships as described above). In this example, the bitmap comprises an array of pixels or voxels, each associated with a print addressable area of a layer of build material to be processed to form a layer in additive manufacturing. The bitmap may for example comprise a binary bitmap, in which pixels/voxels which are associated with a region which is to be solidified are associated with a first value (e.g. 255) and pixels/voxels which are associated with a region which is to be left unsolidified are associated with a second value (e.g. 0).

[0084] In block 704, the pixels of the convolved map which correspond to a particular object portion are identified, for example as described in relation to block 404. For example, a contiguous region in the initial (unconvolved) bitmap having pixels with the first value may be identified. It may be noted that in some examples, the same object may be associated with more than one contiguous region, and these object portions may be considered together or separately.

[0085] In block 706, a mean value of the identified pixel values for that object portion is determined, for example as set out in relation to block 406. In other examples, a different representative value may be used rather than the mean, such as the median.

[0086] In block 708, the mean value then provides an input convolution value into a look-up table which embodies a conversion relationship, for example the relationship shown in Figure 6 resulting in a cooling agent amount scaling value for that object portion. In other examples, the conversion relationship may be embodied as an algorithm, equation or the like.

[0087] In block 710, the cooling agent scaling value is applied to the baseline cooling agent coverage amount for the layer. While a scaling value is used in this example, in other examples a correction to the cooling agent amount may be determined in some other way, for example as an absolute value.

[0088] In block 712, the scaled cooling agent coverage amount is associated with pixels of the bitmap around the object portion. For example, it may be that cooling agent is to be applied within a perimeter around the surface of an object, wherein the perimeter is defined by a distance extending from a normal to the surface of an object, the distance having a predetermined value. The pixels/voxels within this perimeter may be associated with the scaled cooling agent amount. In other examples, the extent to which a region associated with the particular object is defined in some other way. For example, the cooling agent amount may be applied within a boundary box surrounding the object portion, or until a midway dividing line marking the midway between the object and the nearest other object.

[0089] For example, given the relationship in Figure 6, if the convolution value for an object is 31 , which corresponds to the nominal value, then the nominal or baseline coverage amount of cooling agent is used. For example, this may be a contone level of 30. If however the convolution value for an object is 21 , then an increased coverage amount of cooling agent is applied around that object, for example around 1.3 times the amount, or a contone level of around 40. For any convolution value above 41 , then the amount of cooling agent is reduced to around 70% of the nominal amount, but there is no further reduction in this case, to ensure that the build material does not cake as described above. [0090] The method may then loop back to block 704 if there are any other object portions to be generated in the layer.

[0091] In block 714, object generation instructions for the slice are determined based on the (original) bitmap and the associated cooling agent amounts. For example, region(s) of the slice corresponding to the region(s) of the layer which are intended to solidify may be associated with fusing agent amounts, the pixels surrounding the region(s) having associated with cooling agent amounts, which may have been scaled so as to be different from the baseline amount, in block 712. Determining object generation instructions may comprise carrying out a halftoning operation or the like, to determine where print agent drops should be placed in order to provide the specified coverages.

[0092] The method may then loop back to block 702 for another slice of the virtual fabrication chamber.

[0093] In block 716, object generation is carried out. In some examples, this may be carried out in a time frame which at least partially overlaps with the timeframe for carrying out blocks 702 to 714. For example, one or several slices may be processed to determine object generation instructions, and generation of the corresponding layers may commence while object generation instructions of subsequent slices/layers are being generated.

[0094] Generating the object(s) may comprise forming a layer of built material, applying print agents according to the object generation instructions for that layer and providing energy, for example heat, to the layer. This process may be repeated, layer by layer, until the full fabrication chamber is generated, based on the virtual fabrication chamber.

[0095] Figure 8 is an example of apparatus 800, which may be used in some additive manufacturing operations. The apparatus 800 comprises processing circuitry 802, the processing circuitry 802 comprising a print instruction module 804.

[0096] In this example, in use of the apparatus 800, the print instruction module 804 determines a distribution of at least one print agent to be applied to a layer of build material in a layer by layer additive manufacturing process to generate an object. The print instruction module 804 comprises a model analysis module 806 and a cooling agent module 808.

[0097] In this example, in use of the apparatus 800, the model analysis module 806 analyses the intended content of a layer in object generation and determines an indication of the amount of energy to be received by the or each object portion in the layer which is reflected from build material during object generation (the ‘re-radiated’ energy). For example, this may comprise applying a convolution kernel to a bitmap indicative of the intended content of the layer, the convolution kernel modelling the reflection of radiation from the build material and the re-reflection of energy back to the build material from any reflective apparatus features such as a reflector and/or concentrator of the energy source, as described above. In some examples, the model analysis module 806 may derive a value representative of the amount of energy received by the or each object portion, e.g. a mean of the value of the pixels of a convolved bit map associated with an object portion. In some examples, the model analysis module 806 may carry out blocks 702 to 706 of Figure 7.

[0098] In this example, in use of the apparatus 800, the cooling agent module 808 determines an amount (e.g. a coverage amount) of cooling agent to be applied when generating each object portion in the layer based on the indication of the amount of energy for that object portion. In some examples, there may be a predetermined relationship between the indication of the amount of energy and the amount of cooling agent to use in association with the object portion. For example, this may be the relationship determined according to the methods set out in relation to Figures 1 to 6 above. For example, this may comprise reducing an amount of cooling agent relative to a baseline cooling agent coverage when the re-radiation effect is less than a baseline amount and increasing an amount of cooling agent relative to the baseline cooling agent coverage when the re-radiation effect is greater than the baseline amount. The relationship may specify a correction, for example a correction or scaling factor, to be applied to a baseline amount of fusing agent. In some examples, the cooling agent module 808 may carry out blocks 708 to 712 of Figure 7.

[0099] The cooling agent module may use a conversion, for example as described above with particular reference to Figure 6 and Figure 7.

[00100] The print instruction module 804 may also determine amounts of fusing agent which are to be placed in order to cause solidification of parts of a layer of build material corresponding to the object portions to be formed in that layer (for example as described in relation to block 714 above). For example, this may comprise analysing object model data, which may for example comprise part of a virtual fabrication chamber as described above. In some examples, slices of a virtual fabrication chamber modelling objects to be generated (which in this example may be any objects, and may comprise objects which are different to the calibration object and the filler objects described above) may be analysed. For example, a fusing agent and/or cooling agent amounts may be associated with each of a plurality of pixels or voxels modelling the slice. The amount of fusing agent may be based on a predicted temperature within the layer. For example, a predicted heat map of the layer may be generated and instructions to apply fusing agent may be determined such that the fusing agent coverage specified may be reduced in regions of the layer which are predicted to provide ‘hotspots’ compared to regions of the layer which are expected to be cooler. For example, such hotspots may form in the centre of bulky objects. Instructions to apply cooling agent to a region surrounding each object, wherein the amount of cooling agent may be determined based on an indication of the amount of re-radiated energy received by that object portion, as determined by the model analysis module 806 and the cooling agent module 808.

[00101] Figure 9 shows an example of an apparatus 900 comprising the processing circuitry 802 of the apparatus 800 of Figure 8, including the modules thereof. In addition, the apparatus 900 further comprises additive manufacturing apparatus 902 which may be used to generate objects using additive manufacturing. The additive manufacturing apparatus 902 comprises an array of fusing energy sources 904 and a controller 906, although in other examples a single fusing energy source may be provided. In use, the fusing energy sources 904 irradiate a print bed 908 (which may in practice comprise a removable component of the apparatus 902).

[00102] The additive manufacturing apparatus 902 may generate objects in a layer-wise manner by selectively solidifying portions of layers of build material formed on the print bed 908. The selective solidification may in some examples be achieved by selectively applying print agents, for example through use of ‘inkjet’ liquid distribution technologies, and applying energy, for example heat, to each layer using the plurality of fusing energy modules. In some examples, object model data modelling object(s) to be generated may be received and control instructions determined as to where to print agent on a layer of build material in order to generate a layer of the object(s). In some examples, the regions which comprise build material which is intended to fuse are determined, at least in part, by reference to print instructions generated by the print instruction module 804. Such print instructions may be derived based on object model data representing at least a portion of an object to be generated by an additive manufacturing apparatus by fusing build material. The object model data may for example comprise a Computer Aided Design (CAD) model, and/or may for example comprise at least one STereoLithographic (STL) data file.

[00103] In use of the apparatus 900, energy may be provided by the plurality of fusing energy sources 904 to cause the build material to which fusing agent has been applied to fuse.

[00104] The additive manufacturing apparatus 902 may comprise additional components not shown herein, for example a fabrication chamber, at least one print head for distributing print agents, a build material distribution system for providing layers of build material, carriages for sweeping the fusing energy modules 904 across the print bed 908 and the like.

[00105] The controller 906 may control aspects of the additive manufacturing operation. For example, the controller 906 may control the formation of layers of build material, the energy sources 904, the action of a print head to provide print agents and the like. The controller 906 may control additional energy sources, for example build material warming energy modules and the like.

[00106] The apparatus 800, 900 of Figure 8 and/or Figure 9 may, in some examples, carry out at least one of the blocks of Figure 1 , Figure 4 and/or Figure 7. In some examples, therefore, the apparatus 800, 900 may derive the relationship between cooling agent coverage amounts and indications of an amount of reflected energy or radiation received by objects in additive manufacturing.

[00107] Figure 10 shows an example of a tangible machine readable medium 1000 in association with a processor 1002. The machine readable medium 1000 stores instructions 1004 which, when executed by the processor 1002 cause the processor to carry out actions.

[00108] In this example, the instructions 1004 comprise instructions 1006 to cause the processor 1002 to obtain a measurement of a physical property of each of a plurality of calibration objects generated using additive manufacturing in a plurality of layers in which other objects are generated. During object generation, the other objects provide a predetermined packing density (i.e. a predetermined filled proportion of the plurality of layers), and the objects were generated using a predetermined coverage amount of a cooling agent. The measurements comprise a first set of measurements of calibration objects generated with different packing density of the layers and a second set of measurements of calibration objects generated with different coverage amounts. The processor may determine the physical property, for example a weight or some other proxy indicative of a degree of fusion, by controlling a measurement of the objects, and acquiring the measurement as part of that process, or by retrieving or receiving the data from a memory, or over a network, or the like.

[00109] The instructions 1004 further comprise instructions 1008 to cause the processor 1002 to characterise a relationship between cooling agent coverage amounts and an indication of an amount of reflected radiation (i.e. radiation which has been reflected at least once from the print bed) received by the calibration objects, based on the first and second set of measurements. For example, the cooling agent amounts may be determined so that, in a subsequent build operation, objects may tend to have or be closer to an intended object weight, and/or to increase the consistency of weights of generated objects.

[00110] In some examples, the instructions 1004 further or alternatively comprise instructions to cause the processor to generate an indication of the amount of reflected radiation by convolving a bitmap indicative of a content of a layer of build material comprising at least part of (e.g. a layer of) the calibration object and the other objects with a convolution kernel, and determine, from the convolved bitmap, a value indicative of the amount of reflected radiation received by each calibration object. For example, this may comprise carrying out the method described above in relation to Figure 4.

[00111] In some examples, the instructions 1004 further or alternatively comprise instructions to cause the processor to derive corrected print agent amounts for generating object(s) using additive manufacturing by convolving a bitmap indicative of a content of a layer of build material comprising a portion of the object with a convolution kernel, determining, from the convolved bitmap, a value indicative of the amount of reflected radiation to be received by the object portion(s) during object generation; and determining the corrected print agent amount using the value and the characterised relationship. For example, this may comprise carrying out the method described above in relation to blocks 702 to 712 of Figure 7 and/or acting as part of the model analysis module 806 and/or cooling agent module 808 of the apparatus 800, 900 of Figure 8 or 9. In some examples, such instructions may be provided separately to the instructions 1006 and 1008, for example on a different machine readable medium, and may for example be carried out based on a relationship which has been previously characterised.

[00112] In some examples, the instructions 1004 further or alternatively comprise instructions to cause the processor to determine object generation instructions (for example, as described in relation to block 714 of Figure 7) and/or to control an additive manufacturing apparatus to generate an object, for example, as described in relation to block 716 of Figure 7. The generated objects may comprise the calibration object, the other objects and/or a subsequently generated object.

[00113] Examples in the present disclosure can be provided as methods, systems or machine-readable instructions, such as any combination of software, hardware, firmware or the like. Such machine-readable instructions may be included on a computer readable storage medium (including but not limited to disc storage, CD-ROM, optical storage, etc.) having computer readable program codes therein or thereon.

[00114] The present disclosure is described with reference to flow charts and/or block diagrams of the method, devices and systems according to examples of the present disclosure. Although the flow diagrams described above show a specific order of execution, the order of execution may differ from that which is depicted. Blocks described in relation to one flow chart may be combined with those of another flow chart. It shall be understood that at least some blocks in the flow charts and/or block diagrams, as well as combinations of the blocks in the flow charts and/or block diagrams can be realized by machine readable instructions.

[00115] The machine-readable instructions may, for example, be executed by a general purpose computer, a special purpose computer, an embedded processor or processors of other programmable data processing devices to realize the functions described in the description and diagrams. In particular, a processor or processing apparatus may execute the machine-readable instructions. Thus, functional modules of the apparatus and devices (such as the processing circuitry 802, print instruction module 804, model analysis module 806, cooling agent module 808 and/or the controller 906) may be implemented by a processor executing machine readable instructions stored in a memory, or a processor operating in accordance with instructions embedded in logic circuitry. The term ‘processor’ is to be interpreted broadly to include a CPU, processing unit, ASIC, logic unit, or programmable gate array etc. The methods and functional modules may all be performed by a single processor or divided amongst several processors.

[00116] Such machine-readable instructions may also be stored in a computer readable storage that can guide the computer or other programmable data processing devices to operate in a specific mode. [00117] Such machine-readable instructions may also be loaded onto a computer or other programmable data processing devices, so that the computer or other programmable data processing devices perform a series of operations to produce computer-implemented processing, thus the instructions executed on the computer or other programmable devices realize functions specified by block(s) in the flow charts and/or block diagrams.

[00118] Further, the teachings herein may be implemented in the form of a computer software product, the computer software product being stored in a storage medium and comprising a plurality of instructions for making a computer device implement the methods recited in the examples of the present disclosure.

[00119] While the method, apparatus and related aspects have been described with reference to certain examples, various modifications, changes, omissions, and substitutions can be made without departing from the spirit of the present disclosure. It is intended, therefore, that the method, apparatus and related aspects be limited only by the scope of the following claims and their equivalents. It should be noted that the above-mentioned examples illustrate rather than limit what is described herein, and that those skilled in the art will be able to design many alternative implementations without departing from the scope of the appended claims.

[00120] The word “comprising” does not exclude the presence of elements other than those listed in a claim, “a” or “an” does not exclude a plurality, and a single processor or other unit may fulfil the functions of several units recited in the claims.

[00121] The features of any dependent claim may be combined with the features of any of the independent claims or other dependent claims.

Claims

1. A method comprising: generating a plurality of calibration objects using additive manufacturing in a plurality of layers in which other objects are generated, wherein the other objects fill a predetermined proportion of the layers, and wherein the objects are generated using a predetermined coverage amount of a cooling agent; obtaining, by processing circuitry, a measurement of a physical property of each calibration object, the measurements comprising: a first set of measurements of calibration objects generated with different filled proportions of the layers, and a second set of measurements of calibration objects generated with different coverage amounts; and characterising, by processing circuitry and based on the first and second set of measurements, a relationship between cooling agent coverage amounts and an indication of an amount of radiation received by the calibration objects which is reflected from build material during object generation, wherein the relationship is for use in generating subsequent objects.

2. A method according to claim 1 comprising: generating the calibration objects and the other objects in a plurality of calibration batches, wherein the method comprises: generating a first plurality of calibration batches associated with a first predetermined coverage amount and different filled proportions; and generating a second plurality of calibration batches associated with a second predetermined coverage amount and different filled proportions.

3. A method according to claim 1 in which characterising the relationship comprises: determining a relationship between the measured object property and a cooling agent amount.

4. A method according to claim 1 wherein characterising the relationship comprises: determining, for each calibration object and by processing circuitry, an indication of the amount of radiation received by that object that has been reflected from build material during object generation.

5. A method according to claim 4 wherein characterising the relationship comprises applying a convolution kernel to a bitmap indicative of a content of a layer of build material comprising at least part of the calibration object and the other objects; and determining the indication of the amount of radiation received by that object from the values of pixels corresponding to the calibration object in the convolved bitmap.

6. A method according to claim 1 comprising: receiving a bitmap indicative of a content of a layer of build material comprising at least one object portion to be generated; determining, for each object, an indication of the amount of radiation to be received by the or each object portion which is reflected from build material during object generation; and determining, from the relationship, a cooling agent coverage amount for use in association with the or each object portion.

7. A method according to claim 1 wherein the other objects are uniformly distributed over the layers.

8. A method according to claim 1 wherein the other objects have a consistent cross section in the plane of the print bed.

9. A method according to claim 1 in which the measured physical property is indicative of a degree of fusion of the calibration object.

10. Apparatus comprising processing circuitry, the processing circuitry comprising: a print instruction module to determine a distribution of at least one print agent to be applied to a layer of build material in a layer by layer additive manufacturing process to generate an object, the print instruction module comprising: a model analysis module, to analyse the intended content of a layer in object generation and determine an indication of the amount of energy to be received by the or each object portion in the layer which is reflected from build material during object generation; and a cooling agent module to determine an amount of cooling agent to be applied when generating each object portion in the layer based on the indication of the amount of energy for that object portion.

11. Apparatus according to claim 10, wherein: the model analysis module is to: apply a convolution kernel to a bitmap indicative of the intended content of the layer, the convolution kernel modelling the reflection of energy during object generation and determine a value indicative of the amount of radiation received by the or each object portion; and the cooling agent module is to use the value to determine a correction to a baseline cooling agent amount to be applied in association with the or each object portion.

12. Apparatus according to claim 10 wherein the print instruction module is to generate object generation instructions using the amount of cooling agent determined by the cooling agent module; and the apparatus further comprises: object generation apparatus, to generate the object based on the object generation instructions.

13. A machine-readable medium comprising instructions which, when executed by a processor, cause the processor to: obtain a measurement of a physical property of each of a plurality of calibration objects generated using additive manufacturing in a plurality of layers in which other objects are generated, wherein the other objects provided a predetermined packing density, and wherein the objects were generated using a predetermined coverage amount of a cooling agent, the measurements comprising: a first set of measurements of calibration objects generated with different packing densities of the layers, and a second set of measurements of calibration objects generated with different coverage amounts; and based on the first and second set of measurements, characterise a relationship between cooling agent coverage amounts and an indication of an amount of reflected radiation received by the calibration objects.

14. The machine-readable medium of claim 13, further comprising instructions to generate an indication of the amount of reflected radiation by: convolving a bitmap indicative of a content of a layer of build material comprising at least part of the calibration object and the other objects with a convolution kernel; determine, from the convolved bitmap, a value indicative of the amount of reflected radiation received by each calibration object.

15. The machine-readable medium of claim 13, further comprising instructions to cause the processor to derive corrected print agent amounts for generating objects using additive manufacturing by: convolving a bitmap indicative of a content of a layer of build material comprising a portion of the object with a convolution kernel; determining, from the convolved bitmap, a value indicative of the amount of reflected radiation to be received by the object portion; and determining the corrected print agent amount using the value and the characterised relationship.