Transcript for:
Online Training Session: Finite Element Method and CAE - Day One

welcome to on uh to the online training session day one in this I'm going to cover theory of FM and CA the reference book uh the reference of this Theory presentation is taken from uh practical aspects of finite element simulation uh student guide uh uh to download that free copy of student guide you can visit to HTTP double. columns www.al university.com back/ student- gu- request back slash so at the end of the uh session I'm going to show you how to apply for the free copy of student Edition this book where from where I have taken the theory in this training uh series I am going to cover in the as this is the first day I'm going to cover theory of FM on second day I'm going to take 1D machine on day three 2D machine how to create 2D machine or how to generate 2D elements on Mid surface how to edit the them then on day four 3D machine then on day five I'm going to show you the different analysis and postprocessing in day six we are going to discuss topology of optimization day seven I'm going to show you topography optimization on day eight I'm going to show you size optimization on day nine shape optimization and in day 10 composite optimization so this trading series covers hyper Mish complete pre-processing radios and op struct to solvers for linear nonlinear thermal and optimization analysis and and pro processing which includes hyperview so the topic for the today's training session is first we are going to discuss introduction to meshing where the topics will be why do we carry out mching types of element how to decide the element type for a particular model can we solve the same problem using 1D 2D and 3D elements how to decide element length how to start mhing mhing in critical areas then 1D mhing when to use 1D elements types of 1D element then theory of 2D machine in this I'm going to cover when to use 2D elements family of 2D elements that is the types of 2D element how not to mesh what are the different industry standards uh which uh which you have to use while doing the Ming so what are the different things which industry people follow while they are they do mhing then 3D mhing when to use 3D elements degree of freedom for solid elements there there is a slight change in degree of freedom for solid elements then how not to mesh for 3D machine then elements quality check General element quality check particularly for 2D qualities 2D machine or 2D elements quality check other checks for 2D machine quality checks for Tetra machine that is 3D element other checks for Tetra machine brick mesh quality checks again this is a 3D element other checks for Brick mching then lastly we are going to discuss material and property information in which material classifications and material properties so these are the topics which we are going to discuss they are very important to start uh with any CA software so basically CA software has their own engineering knowledge behind them uh behind them so to understand the basic principles of these CA software how they work and what is the basic uh engineering required to understand the behavior of NCA software this Theory portion is very very important so let's start with introduction to mching so the first question arises before entering to anything or before going for CAE the first question arises is why do we carry out machine so the basic idea of FAA is to make calculation at only limited finite number of points and then interpolate the results for the entire domain it may be surface or volume any continuous object has infinite degrees of freedom and it's just not possible to solve the problem in this format finite element method reduces the degree of freedom from infinite to finite with the help of discretization or mesing that is nodes and elements so as in this problem you can see this particular Circle having infinite points and if we calculate 6 degrees of freedom per point so the total equation to be solved is infinite that is not possible but if we discretize this with say eight points like this so at each node we have six degrees of freedom so the total equations become 48 that is quite possible to solve so in this way that is why it is required uh to do FAA so to discretize the degrees of freedom from infinite to finite it is not restricted to number of Freedom eight oh sorry number of nodes to eight it may be 10 it may be four but it's just an example then we move to the next that is types of element so there are different types of elements available starting from 1D we Define 1D as one of the dimension is very very large in in comparison to other two in 1D the element shape is line additional data from the user required is remaining two Dimensions that is area of cross-section then the element types available for 1D element is Rod bar beam P exis symmetric shell Etc and practical applications are long long sha beam pin joints connection elements Etc then 2D elements the 2D or the 2D can be defined as two of the dimensions are very very large in comparison to the third one element shape for 2D machine is called and Tria sorry additional data from the user uh which is required is remaining Dimension that is thickness element type thin shell PL membrane PL stress plain strain exis symmetric solids Etc the practical application of 2D elements or 2D machine is sheet metal Parts plastic components like instrument panels Etc then three 3D elements for a 3D elements all the dimensions are comparable the element shape is Tetra Penta hexa elements or pyramids additional data from the user uh there is no additional data required from the user then element type is solid practical application is transmission casing engine block crankshaft Etc there are some other types of elements are also available like Mass Elements which is known as Point element concentrated mass at the center of the gravity of the component so to put a mass or to put some additional weight to your component you can use that then spring elements translational and rotational degrees of freedom that is translational and rotational stiffness to simulate that we can use spring element then to simulate damping coefficient we can use damper elements to simulate Gap or Gap distances and stiffness friction we use Gap elements to transfer the forces or constraint from one B location to another we use rigid elements like rv2 rb3 and then to simulate welding we use weld elements so these are the different types of elements available then the next question arises is how to decide the element type say for a particular component how we can decide what type of element should we use so the element type selection can broadly be classified as geometry size and shape type of analysis time allotted for the project so first is geometry size and shape we will discuss one by one all so for an analysis the software needs all the three dimensions defined it cannot make calculation unless the geometry is defined completely by mesing using nodes and elements so the geometry can be categorized as 1D 2D or 3D based on the dominant dimensions and then the type of the element is selected accordingly say for example 1D element used for geometries having one of the dimension that is very very large in comparison to other two if you see such type of geometry then then only you can use one elements the shape of the 1D element is nine when the element is created by connecting two nodes the software knows about only one out of the three dimensions the remaining two Dimensions that is the area of the cross-section must be defined by the user as additional input data and assigned to the respective elements practical example is long sha Rod beam column spot welding voltage joints pin joints bearing modeling Etc so as you can see in this model the length is very very large as compared to it cross-section that's why we can use this for 1D element then for 2D element used when two of the dimensions are very very large in comparison to the third one so in some in any of the geometry like sheet metal components where the thickness is very very less as compared to other two dimensions then we use 2D elements to simulate that kinds of problems in FAA of CA so 2D machine is carried out on a mid surface of the Part 2D elements are planer just like paper by creating 2D elements the software knows two out of the three required Dimensions the third dimension that is thickness has to be provided by the user as an additional input data then a question may arise why to create 2D uh why is 2D machine carried out on a mid surface you can use any of these surface either the top one or the bottom one why we have to create a mid surface and then do the mching at the then do the 2D maching on it so mathematically the element thickness specified by the user is assigned half on the element top and half on the element bottom side hence in order to represent the geometry appropriately it is necessary to extract the mid surface and then mesh on the mid surface so in Practical example all sheet metal Parts plastic components like instrument panel Etc in general 2D machine is used for the parts having width to the thickness ratio greater than 20 if the width to thickness ratio is greater than 20 we uses 2dm elements so as you can see in this component the two of the dimensions are very very large as compared to the third one that is the thickness so we Define a mid surface and in this mid surface we are going to do the machine so that half of the thickness will go up half the bottom so the plate simulation will be proper the limitations then we discuss about the limitations of mid surface and 2D machine the 2D machine would lead to a higher approximation if used for variable part thickness also surfaces are not planner and have different features on two sides like if these are at the Top If if it has a different features and then at the bottom of the face if it has a different feature then it is also an approximation because if you generate a mid surface is we cannot create two different features at the top and the bottom then we move towards the next that is 3D element 3D elements are used when all the three dimensions are comparable as you can see in this so if all the three dimensions are comparable so there is no input or no additional data required from the user so practical examples of Trans are transmission casing clutch housing engine block connecting rod crankshaft Etc as you can see the model which is meshed using 3D elements then the second point is based on type of analysis so structural and fatig Analysis we use quad hex elements are preferred over Tri as tetras and pentas because they give the better result then in Crash and nonlinear analysis priority to mesh flow lines and brick elements over tetrodon elements so brick or quad elements are uh preferred over Tria or tetron element for crash and nonlinear analysis for mold flow analysis triangular elements are preferred over quadrilateral element for dynamic analysis when the geometry is borderline between the CL classification of 2D and 3D geometry 2D shell elements are preferred over 3D this is because shell elements being less stiffer captur the mode shape accurately and with a fewer number of nodes and elements then the last point to decide the type of element is time allotted for the project when time is not constrained the appropriate selection of element mesh flow line and a good mesh quality is recommended sometimes due to very tight deadline the analyst is forced to submit the report quickly for such situations automatic or bch Machine Tools could be used instead of time consuming but uh structured and good quality providing methods for 3D machine tetras are preferred over hexas because it they can we can do mesing much faster using Tetra elements if the Assembly of several component is involved then only the critical parts are meshed appropriately other parts are either course mesh or represented approximately by Oney beam spring concentrated Mass Etc then the next is can we solve same problem using 1D 2D and 3D elements it is not possible to use 3D elements for long slender beams 1D geometry or sheet metal Parts 2D geometry and 2D shell elements for representing big casing parts for same geometry the same geometry could be modeled using 1D 2D 3D elements what matter is the number of elements and no degree of Freedom the accuracy of results and the time consumed in the analysis say for example consider a can Beam with Dimensions 250 20x 5 mm that is subjected to 35 Newton Force if we calculate using 1D element the number of nodes are two and number of element is one if we use two 2D shell machine the number of nodes is 99 number of elements are 800 so total degree of freedom is 54 54 if we use 3D elements for the same problem then total number of nodes is 1,448 number of elements goes to 9,569 and total degrees of freedom goes to 52344 but there is no changes in the the results 1D element is giving the same result as 3D but as the number of nodes and elements increases the solver time also increases so depending on the requirement of as per the geometry and as per the availability or as per this way of analysis you want to perform decide the type of element then the next question arises is how to decide element length we can do the machine of a particular component with element size 2 3 4 5 6 7 4.5 whatsoever we can pick any length so how to decide which length is the best so there these the element length is decided as based on previous experience with a similar type of problem successfully correlation with experimental result results then type of analysis linear static analysis could could be easily carried out quickly with a large number of nodes and elements but crash nonlinear cfd or dynamic analysis takes a lot of time keeping control on the number of nodes and elements is necessary then Hardware configuration that is very very important and graphic card capacity of the available computer an experienced CA engineer knows the limit of nodes that can satisfactory handled with the given Hardware configuration suppose you are a part of newly formed CA group no clear guidelines are available and there is no experienced person in the group in the first run then you can do this like in the first run accept the default element length mesh it with the basic rules then run the analysis and observe the high stress region rematch the localized area of the high stress with smaller element length and solve it again compare the difference in the original and the new results continue the process until the convergence is achieved that is 5 to 10% difference in strain energy and maximum stress value between the two different element length when this is achieved you can pick the earlier one so say for example you are mesing a certain component critical Zone with elements length as eight you get a stress as 100 then you lower down the element size to say five and you get a stress as one1 so there is no much difference in the two stress results so you can pick 100 and the element size is 10 and carry out the machine then the next question arises is how to start the machine from where exactly we should start the machine so to start the machine first point is to spend a sufficient amount of time studying the geometry a common observation is that c ER starts meshing immediately without properly understanding the geometry and paying attention to all the requirements and instruction provided observing the geometry SE several times and thinking about it for all angle is strongly suggested mental visualization of step is the first step in the right direction of creating the good machine then time estimation nowadays the trend is towards the client or boss specifying the estimate time for a given job to the service provider or subordinates sometime it is decided based on Mutual understanding a Time estimation is very relative and one can find a lot of difference in estimation by different engineer as much as two to three times usually a less experienced person will estimate more time also if someone is handling the job for the first time then he shall he or she will require more time if similar kind of job is given to the same engineer again and again the M time will reduce drastically geometry check generally cat data is provided in iges format geometry cleanup is an integral part of mching activity CA engineer should at least have the basic knowledge of CAD before starting the job the geometry should be carefully checked for free edges scale lines duplicate surfaces small fillets small holes bids intersection of Parts assembly of the component if suppressing fillet small holes bid or the generation of mid surface is required for missing then why isn't the cad data provided in the way needed for CA by the cad Engineers itself the answer is yes theoretically that would be an idle situation but practically everyone works with a very tight schedule and a Target dates cat data is generated keeping in mind the final drawing to be released for manufacturing the same CAD model is provided simultaneously to the tools jig pictures manufacturers vendors purchase engineers and CA Engineers as well the simplification required for fa is understood better by CA engineer than a cad engineer hyperworks software provides special tools for geometry cleanup and simplifications which are usually much faster than CAD software many times for liated geometry surfacing operation fails in cat software and it could not could be easily handled by CA engineer by avoiding the geometry and generating the mesh using manual or Special meshing Operations so that's why uh the cad is not designed like that then the fourth point is symmetric check to start the machine the fourth point is symmetric check we can check for the complete part symmetry mesing only a quarter of the plate and reflecting it twice is advisable as you can see this quarter plate if you reflect it once you will get a half twice you will get a full plate so you you can also check for the Symmetry then subpart symmetry repetition of features and the copy paste command maching the highlighted 22.5 Dee portion and then using reflection and rotation would lead to a faster mesh as well as same structure of elements and nodes around the critical area that is holes so we can cut a section of some some angle and we can use the reflect rotate command to do the mhing do the do the same mhing everywhere around the circular as you can see in this rim for this particular Rim then selection of type of elements in real life we rarely use only one type of element it is usually a combination of different types of element that is 1D 2D and 3D and others in the below figure the handle of the bucket is model with 1D beam or 1D element the bucket body uses shell 2D elements and the connection between the handle and the bucket body through rb2 rigid element so these are the different types of elements a CA engineer uses to model a real life problem then type of meshing if it's a geometry based mesing the mesh is associated to the geometry if the geometry is modified the mesh will also get get updated accordingly automatically the boundary condition could be applied on the geometry like surface or Ed Etc and Fe based mesh is not associative the boundary conditions are applied on elements and nodes only so you have to decide the type of machine you are going to perform geometric based or Fe based then how to do meshing in critical areas first of all we will try to understand what are the critical areas critical areas are the location where high stress locations will occur dense meshing and structured mesh no Trias no pentas is recommended in these regions areas away from the critical areas are called general areas geometry specific ation and course mesh in general areas are recommend to reduce the total doof at solution time then how would I know about the critical area before carrying out an analysis after going through a previous analysis of similar part carried out by your colleague or a senior in group one can get a fairly good idea about the probable location of High stresses but Suppose there is no past record and you are going uh you are doing it for the first time then run the analysis with a reasonable element length and observe the result high stress regions are critical and could be remed with a smaller element length in the second run rules for modeling holes and fillets so critical areas are circles so minimum 12 elements around the hole all for general areas four to six element will also work for critical areas the fillets minimum three elements on the fillet you can for the general areas you can suppress the small fillets or one element for large fillet will work also mesh transition techniques and flow lines there is different mesh transition techniques like 1 to 3 uh 2 to 4 1 to2 or like 1 into 2x2 then we will start with 1D mhing so so now the question arises when to use 1D element 1D elements are used when one of the DI di mention is very very large in comparison to other two the element shape is line the representation of this it looks like line additional data from the user required are the remaining two Dimensions that is the cross-sectional area the element type is Rod bar beam pipes exis symmetric shells Etc practical applications are long shaft beams pin joints connection elements as you can see in this then we will see the different types of 1D element so we'll start with Rod for Rod element it supports uh it supports only tension and compression and torque for some softwares like in hyperox it's suppos torque also so ux or UR stands for translation R stands for rotation translation about particular AIS rotation about that particular exis the Practical example may be tension compression members like trusses shaft subjected to torque connection of the elements then bar elements it supports all the six degrees of freedom that is translation along XY Z AIS and rotation about XY Z AIS it is applicable only for symmetric cross-sections the Practical examples can be stated as Sha subjected to multi exual loading bolted welded joints connection elements Etc same as bar beam elements uh beam 1D elements same as bar but also some more unsymmetric cross-sections Shear centers and War Pages then same as bar the practical application but for and plus additional to that for an symmetric crosssections also then pipe elements same as beam except it has internal nonzero di that is it's a hollow then practical application is Ping system structural analysis then ex symmetric shells they only supports Z AIS of symmetry X as radial for object symmetric about the axis of rotation and subjected to exis symmetric boundary condition thin chel press vels cylindrical conical objects Etc then we discuss about uh 2D machine the basic question is when to use 2D elements 2D elements are used when two of the dimensions are very very large in comparison to the Third Dimension the the element shape for 2D elements are quad and Tria additional data from the user are remaining two remaining Dimension that is the thickness as you can see the thickness which we give as the user defined value then the element types available in hyperox is thin shell plate member plain strain plain stress exis symmetric solids Etc then practical application as sheet metal Parts plastic components like instrument panels Etc then why is 2D maching is carried out on the mid surface quite often the geometry of thin Vault 3D structure as shown in the image below is simplified to a geometric model with lower dimensionality this is typically called as midsurface model the midsurface model is then mesed with 2D elements thus there is no need for a detailed volume Mage as the thickness of the geometry is virtually assigned to the 2D element mathematically the element thickness specified by user is assigned half in plus Z as you can see for this model plus Z Direction and the other half in negative Z Direction element bottom so that's why we create uh midsurface to perform 2D machine then the family of 2D elements or different types of 2D elements available in hyperox is Tria and cord there are further classification for tri and cord is constant strain triangles that is linear triangular elements or linear strain triangles that is parabolic triangular elements they are also known as first order Tri elements and second order Tri elements in second order Tri elements the number of nodes increases so the number of interpolation points increases same the classification is for cord linear cord having four nodes parabolic cords Having Eight nodes then the most important part is how not to M there are different different rules uh uh generated or assigned by the industry that how not to mes which which start with backto back triangles should be avoided two Tri element should not be connected to each other directly as you can see in the circle regions that two triangular element should not be connected back to back as they increase the stiffness triangle elements are more stiffer as compared to cord on a plain surface triangular element should be avoided the reason is same they are stiffer so and they can also deviate the mes low lines so it should be avoided in the planer if it's possible go for the cad elements no mesh transition or constant radius fillets or curvature the mes transition should be carried out on the planer surface so in the fillets there should be no mes transition it should be either on the planer regions because it may or may not come under the high stress region avoid Tri elements on Outer Edge or holes so at the outer edges because there those are the failure regions so avoid here and around the holes also avoid the Triangular elements what is not accepted uh acceptable at professional level is this type of machine which is not structural or flow what is acceptable is this with a proper flow of elements then circular holes should be modeled carefully with a washer 1.5 to two times Di and a minimum of two layers around the hole as you can see so that the uh interpolation will become strong and the results will be more accurate holes should be modeled with an even number of equally spaced element for a better representation of whole geometry and uh smooth mesh flow lines whole should be modeled with a even number of elements like 6 8 12 16 Etc rather than 5 7 9 and 13 to get a perfect whole like structure node should lie properly on the surface with no deviation and no Kinks so as you can see there are some Kings so these nodes are coming out of the surface so this should not be done it should be properly on the surface switch how how you can see that in hyper mesh switch off the element mesh line and observe the Contour in particular at curvature Kings as shown above are not acceptable follow the feature line node should lie exactly on the edge so you have to follow the surface feature lines maximum time that you have to keep in mind that you have to follow the feature lines that instead of zigzag distribution a structured or smooth mesh is recommended node align in a straight line use the smooth option available in hypermesh provided by most of the softwares help in achieving systematic mesh so this type of meshing is not allowed in professional the mesh is perfectly in flow for crash analysis flow of the mesh flow line requirement required so it has to be perfectly in flow for crash analysis rotating cords are not allowed as you can see this type of structures are not allowed for crash is for p analysis constant mesh size by using Trias is preferred due to minimum element length and time step criteria so this type of structure is allowed instead of this one then we will start move towards the 3D machine so the again the question is when to use 3D elements so 3D element should be used when all the dimensions are comparable that is all the three dimensions are comparable then only we use 3D elements as you can see in this that is XY Z all the three are comparable the element shapes available for 3D element is Tetra Penta hexa and pyramid additional data from the user required is nothing because we create the elements in all the three dimension itself here the element types are solid practical application is G GE box engine block crankshaft Etc these types of things which are model using solid elements then the different types of 3D elements or you can say family of 3D element includes Tetra element they are classified again like linear Tetra having four nodes at the corner or a parabolic Tetra having 10 nodes additional nodes at the mid side of their edges of the elemental edges then Penta elements linear Penta having six nodes parabolic Penta having 15 nodes hex or brick mching or brick element Having Eight at the nodes eight nodes at the corner or 20 nodes in the parabolic uh H element then we will come up with uh degree of freedom for solid elements like for 2D thin shell and 1D beam elements suppose supports six degrees of freedom like three trans translational and three rotational that is translation along XY Z and rotation about XY Z AIS but all solid elements have only three translational degrees of freedom no rotational degrees of freedom that is a 10 noed Tetra element has only 10 into 3 that is 30° of freedom why does a solid element have only three translational and no rotational degrees of freedom let's physically interpretation interpretate that consider a piece of paper 2D geometry or long uh say it's a 2d geometry like or a long steel scale that is a 1D geometry it could be easily bent or Twisted that is rotational degrees of freedom but now consider a solid object like Duster or paper weight it could not be subjected to very high bending or tsal stiffness hence solid elements have been formulated with three translational degrees of freedom and no rotational degrees of freedom as you can see in the images then how not to mesh for 3D mhing for first point is mid node should lie exactly on the geometry for a parabolic that is second order Tetra mching thas many CA Engineers prefers to start with linear Tria that is first order instead of parabolic meshing and then convert it into parabolic that is second order in the conversion process mid nodes might not get projected automatically on the curved surface and fillets if so it should be be projected on the corresponding surfaces before conversion to tetras so the midpoint should lie project particularly on the surface as you can see here when the job is split among several Engineers the element length and the overall mesh pattern should be consistent the above job was split among three Engineers due to very short time duration provided by the client the same mesh size and pattern was not followed by the engineer working independently on subject on sub parts of the geometry as you can see here so the mesh element size is different different and different so it should not be like this if you split your geometry and divide it into n number of users to do a same part the splitting portion should be meshed first and the all the uh all the other engineer has to start from that region so the mesh consistency will be there then minimum two elements on the fillet of the Tetra mchine elements at the fillet and curves are pH usually fail as you can see here so two elements one two at the fet so elements at the fillets and curved surfaces usually fails in the Jacobian Distortion elements check the manual adjustment and improving the element quality results in mesh deviation from the geometry and visible Kings this could be avoided by modeling the fillets with two or more elements then for Brick meshing a minimum of two elements across the thickness should be used a single element lead to a poor interpolation and does affect the accuracy of result a minimum of Two element across any thickness is recommended the exception in nvh application where the stresses is not the main criteria but the representation of mass and stiffness with the least degree of freedom is the main criteria use of tetra pentra uh pyramid element while brick missing some clients allow for a few Tetra element during brick mhing also some softwares and Analysis type supports p PID elements like hyperworks uses the Tetra and pyramid elements can make the life of the brick measure tolerable it's good practice to clarify the instruction for the use of these elements from the client as you can see pyramid or Tetra modeling a sheet metal part with 3D element for Sheet Metal a very small thickness Part 2D shell elements are better suited for and recommended it is not like we cannot use 3D M but it will result in very high number of nodes and elements considering the following sheet metal part we will mesh the same part with 3D parabolic Tetra element and 2D C for linear element using the same element length and compare the number of nodes and elements needed so for 3D Tetra M the number of nodes is 1496 and elements is 689 while for the same component the number of nodes is one to one and elements are 100 so you have to decide which component you want to go for 3D and which component you want to go for 2D limitation of 1D element and advantage of 3D machine fillets cutouts and complicated geometry features cannot be represented accurately by 1D element 3D element because of three dimensions can capture all the Minor Details accurately for example considering the following shaft it is very difficult to capture the keyway slot and variable fill it using 1D element instead 3D maching is recommended for such applications then we will move to elements quality check so now uh we'll discuss about elements quality and check so element quality is a subject often talked about and never fully understood the reason for this is complex but is related to the fact that quality is relative and the Solution by definition is approximate in the formulation of finite element a local parametric coordinate system is assumed for each element type and how well the physical coordinate system both element and Global matches the parametric uh dictates element quality below you will see you see some graphic representation representing element quality and you should attempt to follow them however there will be a point of uh diminishing results if you try too hard to get every element within the acceptance criteria your judgment is your only guide in those cases always perform quality check on the meshes you create check with local experts regarding the appropriate values for each element type required by your element checking computer program beware that in this situation correct answer can vary a great deal as Illustrated in the following table where the range between okay and very four is quite wide solid elements uses use the determinant of the Jacobian metrix and compare to the idle value some common element quality measures are detailed BAU the best element 2D Tria is 60 60 60 and all the sides are same okay is right angle Tria and somewhere two sides are equal one is different very poor is like the third one with a 10° or 105° angle same for the quad perfect is perfect idle Square then a r triangular and some other changes in the side or angles and very poor is the 30° angle triangles uh quadrilaterals so we'll discuss some important quality checks like skewness skew in Trias is calculated by defining the minimum angle between the vector from each node to the opposing mid surface and the vector between two adjacent mid sides at each node for the element as you can see here 90° minus the minimum angle found is reported as skewness skew in cord is calculated by finding the minimum angle between two lines joining the opposite mid sides of the element 90° minus the minimum angle found is reported Theus Q check is performed in same Fashion on all the faces of 3D elements then aspect ratio spect ratio in two dimensional element is calculated by dividing the maximum length side of the element uh of an element by the minimum length side of the element the spe ratio check is performed in the same Fashion on all the faces of 3D elements warpage war page in two dimensional element is calculated by splitting a quad in two Trias and finding the angle between the two planes which the Trias form is then found the maximum angle found between the plane is the warpage of the element warpage in threedimensional element is performed in the same Fashion on all the faces of the element then now we will discuss about some detail quality checks by taking 2D quality checks first for 2D quality checks the idle shape for Quad element is square the idle shape for triangular elements is equilateral triangle different quality parameters like skew spect ratio include angles Jacobian stretch Etc are the measures of how far a given element deviates from the idle shape a square means that all the angles are 90° with equal sides while an equilateral triangle has all the angles at 60° with equal sides some of the quality checks are based on angles like q and included angles while other on side ratios and areas like expect and stretch to reduce the solution time elements are mapped to local coordinate system individually for every element at the centroid instead of using a single coordinate system that is the global coordinate system the effectiveness of this transformation is checked by the Jacobian and Distortion ideally all the nodes of the Quad element should lie on the same plane but but because of curvature and complicated geometry profile it is not possible the measure of the out of plane angle is the warpage angle so the warpage angle is the out of plane angle ideally it has to be zero but acceptable value is 10° warpage angle or the wrap angle is not applicable for triangular elements it is defined as the angle between the normal of two planes formed by splitting the cord elements along the diagonal the maximum angle of the two possible angles is reported as warpage angle as you can see so this is the warpage angle and if we extend this this is the warpage angle then spect ratio for 2D elements is the maximum element Edge length divided by minimum element Edge length ideally the value should be one but the acceptable range is should be less than five the ratio as you can see the minimum length maximum length to minimum length ratio should not exceed five then SK idle value of SK is zero but acceptable is less than 45° Q for quadrilateral element is 90° minus the minimum angle between the two lines joining the opposite mid sides of the element Q of the Triangular element is 90° minus the minimum angle between the lines formed each node to the opposite mid side and between the two adjacent mid sides at each node of the element as you can see here Jacobian the idle value of Jacobian is one but acceptable should be greater than 6 in simple terms the Jacobian is a scale factor arising because of transformation of one coordinate system to the other elements are transformed from Global coordinate system to local coordinate system defined at the centroid of every element for faster analysis time then include angle that is very important like Q is based on overall shape of the element and it does not take into account the individual angle of quadrilateral or triangular element included or interior angles check is applied for individual angles for quadrilateral or the quad idle value is 90° but the acceptable range is 45° to 135° for Tria elements idle value is 60° but the acceptable range is 20° to 120° then minimum element length this is very important check for crash analysis that is time step calculation it is also applied in general to check for the minimum feature length captured and the presence of any nonzero length element so if in minimum length if anyone is getting zero length that means the feature is going very very very worst that particular element so we have to remove that then C deviation this helps in determining how well curvatures have been modeled it is defined as the distance between the midpoints of an element Edge to the curved surface it is not applicable for linear elements then then how to improve the quality of poor elements there are different methods to improve the quality of poor elements like manual adjustment this is done by translating the nodes manually or remeshing in the poor mesh region the method consumes lot of time and was the only technique available for years drag node the user has to drag the node of the failing element it works faster and the advantage is that it instantaneously show the effect of dragging the node on all the attached Elements Auto quality improvement programs this is the latest option for quality improvement the user has to submit the mesh for quality improvement and software program runs in the background to improve the elements quality automatically so these are the different options available in hyperworks to improve the poor element the quality of elements then some other checks for 2D machine is element free edges what is free Edge any single quad elements has four edges as you can see here two element in the case above the middle Edge is the shared and is no longer a free Edge you can see here for a real FEA model free Edge should match uh with the geometry Outer Edge that is free edges any additional free Edge are an indication of unconnected nodes so the free Edge should be at the outer edge of the surface any in between free Edge means there is no connectivity or unconnected nodes so it should be avoided you have to be a properly connected otherwise the forces and conate will not get transferred from one portion to another then duplicate element mistakes during the operation like reflect or translate can result in duplicate elements these duplicate elements do not cause any error during the analysis but increases the stiffness of model and result in small displacements and stresses for example consider a simple plate thickness 2 mm subjected to a tensile load assume that due to some mching operation all the elements are duplicate if the analysis is carried out then it will show half the stress and displacement then duplicate nodes operation like copy translate Orient or reflect can result in duplicate nodes at a common Edge in the above image there are duplicate nodes at the interface where the mesh was reflected the duplicate nodes are highlighted in yellow then geometry deviation after the completion of meshing the geometry the mesh and the geometry should be viewed together mesh line option off the mesh should not deviate from the geometry delete free temporary nodes free nodes if not deleted results in rigid body motion when the auto singul option is turned on the software uses a spring element with with a very small stiffness to connect the free uh stiffness to connect the free node with the parent structure this results in warning message during the analysis reum nodes elements property Etc before export operation frequent import export operation could lead to a very large number of number for nodes and elements IDs some software refuses to read the file in the node element node or element IDs are greater than a specific limit this could be avoided by re numbering the nod and elements observe type family and number of elements element summary for complete model the mesh should be checked carefully prior to the export operation as well as after importing it in the ex external solver for element type family number Etc sometimes due to translator program if properties are not defined properly or for non-supportive elements either the element are not exported at all or the family is changed like membrane element converted to the thin shell Etc blot Trace line element free Edge or free phas if any should be deleted check Mass actual mass versus Fe model Mass when a prototype or physical model of a component is available the Fe model Mass should be compared with the actual mass a different a difference means that there is there are missing or additional components or uh improper material or physical properties free free run or dummy linear static analysis before delivering the final mesh to the client a free free run should be performed six rigid modes indicate that all the parts in the assembly are property connected to each other in the case of a single component mching job a linear static analysis with a dummy boundary condition should be carried out to check the mesh quality mesh connectivity request your colleague to check the model due to Contin continually working on the same project your mind tends to take some of the things for granted and there is a possibility of missing some of the points it is is a good practice to get it cross checked by your colleague prior to finally delivery then we'll discuss quality checks for Tetra machine first we will start with T collapse the idle value is one and the acceptable is greater than 0.1 tetraols calculated as H multiplied 1.24 divided by a it is defined as the distance of node from the opposite uh phas divided by area of the phase multiplied by 1.24 other checks are same for the Tetra elements as 2d1 that is Jacobian and uh included angles and everything the other checks for Tetra mesing is quality check for 2D Tri element before converting Trias to Tetra all the quality checks as discussed for the shell element should be applied free Edge conversion from Tria to Tetra is Possible only when there are no free Edge no free Edge indicates the meshes in a enclosing volume T connections a mesh model should not contain any T connections for Tetra mhing as you can see here this is the geometry T connection not acceptable green element should be should not be there in the model no T connection corrected mesh consistent shell normals before converting Trias to tetras the shell normal should be corrected some software do not allow shell to solid conversion unless the normals of all the elements are properly aligned geometry deviation after the completion of meshing the geometry as well as mesh should be viewed together mesh line option off the mesh should be should not deviate from the geometry in the process of quality improvement in particular uh for Distortion Jacobian on curved surface or fillets sometimes not get translated too far away from the geometry and is not acceptable 2D Tri element should be deleted before the final submission it's a common mistake to export 2D shell elements along with the Tria tetra mesh in the final delivery then we'll discuss about brick mesh quality checks the idle shape of the brick element is Q various quality check criteria check how far a given element deviates from the idle shape warpage angle idle value is zero and acceptable should be less than 30° warpage angle is calculated on face of the quadrilateral element of of a hex element it is the angle between the plane that form by splitting the quad element Jacobian idle value is one acceptable should be greater than 0.5 in simple language the Jacobian is a scale factor arising because of transformation of the coordinate system elements are transformed from Global to local coordinates to reduce the solution time aspect ratio idle value is 1 acceptable is less than five aspect ratio is equals to maximum Edge length divided by minimum Edge length cord phas include angle is 45° to 135° Tri pH include angle is 20° to 120° number of percentage of pentas acceptable for a brick maching should be less than 5% other checks for Brick mching free faces a free face check is the most important check for Brick mhing a single brick element has six free faces free faces of the mesh should match with the outer surface skin of the solid part any extra inside faces indicate that either nodes are not connected properly or there are mismatching elements as you can see this is the brick machine free faces at the inner side you can see the free face inside also so it is not connected also converting the free faces to Tetra for complicated geometries checking the internal fa free Phase could consume lot of time a quick shortcut is to convert the free Phase to tetr mesh successfully converion indicate that the brick mesh is okay and there are no internal faces so instead of finding it uh for the complicated geometry the inside free faces you can directly convert that into Tetra if it converts into Tetra that means there are no free edges then the last topic of today's webinar is material and property information so the material classification we will discuss with isotropic ISO means same Tropic means Direction isotropic has the properties independent of direction or axis two independent constant that is e or new they are mostly Metals orthotropic Oro means three Tropic means Direction different properties along three XIs nine independent constant would concrete rolled Metals Etc an isotropic different properties along crystalographic plane 21 independent constant all real life materials are an isotropic only but we simplify them into the category of isotropic or and orthotropic laminates two or more Metals bounded together in layers simplest example is lamination carried out on uh certificates identity cards Etc mainly used for space applications and these days in automobiles the trend is Shifting towards plastic and laminates from metal as you all know that then material properties these are the this is the table which discuss the modulus of elasticity poisons ratio density yield strength and ultimate strength for different materials like Steel Cast Iron rot iron aluminium aluminium alloy brass bronze copper copper Alo magnesium titanium glass Rubber and concrete you may uh be discussing that why mass is in ton and density tons per mm CU for Newton mm unit system so the explanation for this is 1 Newton is equal to 1 kg into 1 m per second Square f = m a so 1 Newton is equal to 1,00 kg multiplied 1 mm by second Square so 1 Newton is equals to 1 T multiplied by 1 mm by second Square hence when force is in is specified in Newton length is in mm Mass must be specified in tons and density in turn per M Cube the different systems for steel is for SI the length unit is meter mass unit is kilogram time unit is second Force unit is Newton for MTS system length is MM mass is time is second force is Newton for mm kg Ms system length is in mm mass is in kilogram time is in milliseconds and force is in kilon so the young modulus will change as for SI system it is 210 e to the power 9 for MTS it's 210 e to the power 3 and for mm k G Ms it's 210 same the value changes for density but the poison ratio Remains the Same at is is it it is a unitless quantity so whenever you enter uh the material properties and software make sure you enter the correct material properties to get a perfect results in this way I have ended the first uh session first day session of this 10day hyperworks training thank you very much for attending this session to apply for the free E copy of the book please visit HTTP www.university.com SL student guide request to improve the online training session please help us by sharing your valuable comments at edus support india. ala.com