SIMULATION in FUSION 360: IMPROVE the STRENGTH of your 3D prints!

[Music] joël the 3d printing nerd designed and tested 3d printing filament rack brackets back over on his channel a couple of weeks ago to see how they perform and then revise the design based on his destructive tests since fusion 360 also has a finite element simulation tool built-in all today show you how you could also optimize your design using these numerical tools and then I'll test which of the brackets can withstand the highest amount of load guten tag everybody I'm Stephan and welcome to CNC kitchen fusion 360 has quite a couple of different simulation tools built in which you can also use to analyze and optimize your own designs in a previous video I already showed you how you can use the topology optimization tool to create an optimal structure for a given design space with the applied constraints and forces in this video I want to show you how you can use two of the other structural simulation modules to analyze your existing design and optimize it with the results to get a part that performs as good as possible Joel has designed these brackets as part of a filament rack you can screw these to the wall and then just add 1 by 2 lumber pieces and have yourself a nice fellowman drag in order to find out how much load they can withstand he built himself a crew test rig and loaded his two different designs printed in different materials until they failed taking a look at his tests showed that his designs failed due to two reasons structural failure due to high stresses and buckling both of these types of failure can be simulated in fusion 360 so we'll take a look at his designs and then optimize the brackets with this information to get an even better performing part in the end this won't be a complete course and finite element analysis but should illustrate you how you can use the given tools to improve your designs before you test them the first time to save design iterations also simulating FDM 3d printed parts is very complex because the material properties are directionally dependent and we are working with mostly parts that are Hollow and fusion 360 is just not made for that so our simulations won't be 100% precise but still should give us an indication where the problematic regions are and if a failure is likely I also know that the load case is not perfect since we load only the front but it is the worst location and for this load case we have test results to correlate against so without further ado let's jump right in Joel was so kind to provide both fusion 360 models of the brackets we have the first one that he used for his first test and the other one which is a little bit reinforced that he used for his second test in order to perform a simulation on this model we need to apply a material to the part and in this material the mechanical properties are stalled since fusion 360 doesn't have a material model for PLA I just used the polystyrene one where the Youngs modulus is quite similar to PLA and should be totally sufficient for our simulation when we enter the simulation environment we have the option between choosing quite a lot of different simulations starting from the static analysis over it the structural buckling analysis all the way back to shape optimization taking a look at the failure modes of Joel's brackets we are at the moment basically only interested in the statical analysis and also the structural buckling analysis because quite a lot of his first brakus just failed due to the stability problem at first I simply apply a false at the location where Joel also has been attaching his strap where he added the Lord through the bracket for the failure tests taking a look at the states are most of his brackets failed around 200 to 400 Newtons so for this analysis I just went with 300 Newtons next we need to apply the constraints the constraints were that the bracket was bolted to the wall so I used fixed supports at the holes so the analysis is always a bit of a simplification because if we only use these boundary conditions we do not really take the wall into consideration and this might affect the stresses in the end so in the first analysis I'll just use this simple constraint and in a second analysis I'll also add a additional constraint which shall represent the wall and we take a look at the stresses of both next we need to measure our part meshing more-or-less divides our parts into smaller chunks and these model chunks can be described mathematically so the solvent the background can calculate all of the displacements and stresses this procedure also called discretization is something that is kind of simplifying our problem and the thing is the final Umesh is the more accurate in the end your results will be but the issue is that the finer your mesh gets the more complex your simulation gets and the more computational time it will take so in the end you basically need to find a way between a really fine mesh and a really coarse mesh which is still giving you quite accurate stresses in the end I'll go into my settings and just slide the slider all the way to the left that we get a kind of fine mesh in the end if you'd be serious with finite element analysis you'd always perform a mesh refinement study to find out if your mesh was even fine enough so just some basic rules that you should have in your head when you're doing finite element analysis if you have sections as we see it right here you should at least have two elements over the thickness of a part and this is really the bare minimum because otherwise the simulation results might be way off additionally in especially the radii you should have at least three elements over an angle of 90 degrees the finer the mesh will be in these locations the more accurate the simulation will be but just as I said before you always need to find a way in between a really fine mesh and a really coarse mesh I think for the analysis we want to do at the moment this mesh density is totally sufficient because we don't want to take a look at the real precise stresses but just the general stress distribution so I won't change anything for the moment here okay so the next thing is basically just running the analysis so there are two options you can either run the simulations in the clouds on the service of autodesk or you can run it locally depending on the size of your simulation and well the the computer you are running your simulation on it might be a good idea to run it in the cloud because this is for bigger models usually faster but if we have a small problem as we have a red here you can still run the simulation locally in your machine and a problem like this will be finished in a couple of seconds or a couple of minutes okay and so this is basically the result of the first analysis and the first plot we see right here is the calculation of a safety factor and the safety factor is more or less just a comparison of the stresses you have in the model to the material values that are in the background for the material the thing is we have used polystyrene and as I've also said in the beginning the material properties of 3d printed parts are not really as a tropic so they differ in different directions so I think taking a look at the safety factor for such a model isn't the best thing so I'll change the solution from safety factor to pharmacist dress the pharmacist dress is a calculated equivalent stress that more or less calculates from a complex multi axial stress state a single stress value that we can use for the comparison to the material values of our part so for PLA if you have been takin a look at the videos I already did about some some brands of material you would know that in printing Direction the material is actually pretty strong and can even withstand stresses up to 50 60 or even 70 mega Pascal's I use a surface probe right here just to check the stresses that we have in one of the worst locations and with a stress of around 50 mega Pascal's we are actually already in the regime where a failure could actually happen so if we want to load our part with the 300 Newtons this might be already enough to produce a failure on our part the other location you see down here at the location of the bolt is more or less an artifact due to our boundary conditions that we have chosen in a real-world example if the bracket would be bolted to the wall this well could not happen since the bracket would be supported and this bending motion wouldn't be realistic so in order to also take the wall into consideration I slightly changed my boundary conditions and I add an additional constraint to the backside of the bracket and use a frictionless support there to more or less mimic the wall to be honest this is also not 100% perfect the bracket is not able anymore to move away from the wall if you would like to model that you would need to use contacts but this is a totally different story and would make the calculation really more complex so I don't want to go in more details about that at the moment again I just solve it wait for a couple of seconds and then we can take a look at the results okay so what you can see is that basically the highest stress concentration down at one of the screws is gone since now we are supporting the bracket at this location but still the location just where we also load the part is still there and also taking a look at the results from Joel's tests this is actually one of the locations where the bracket might be too weak and where in a design optimization we should add more material to get rid of the stress concentration okay so the next thing we want to take a look at is the stability behavior of this bracket as we have seen in Joel's test as some of the brackets he has tested failed due to buckling buckling is a stability problem and can already happen way before the normal material failure of a part since our bracket is also kind of a slender structure we also need to run a buckling analysis and we also already have the evidence from the tests that Joel has performed that this is a real failure mode that will happen in real life so again I just apply a load of 300 Newtons at the location where also he tested his bracket and had the constraints add the screws and I'll also already directly add a frictionless support to the back of the bracket where it is bolted to the wall and again it is necessary to add a mesh to the part that is fine enough that the stiffness of the part can be calculated correctly now it's a little bit different than before because fusion 360 does not really let us solve this calculation locally on our machine so we all of the time have to send it to the cloud I don't know why but that's just how it goes this simulation usually takes a little bit longer to solve than the normal static analysis but still after a couple of minutes you'll get the results back okay and again we are greeted with some pretty pictures of the buckling modes that were calculated the buckling factors which are presented to you next to the legend are more or less a multiplier how big the load would need to be that such a failure can happen interestingly the first mode that we see has a negative buckling value or a negative buckling multiplier and this means that you would need to apply your load in the opposite direction that such failure would be feasible so since we are just loading our part in a positive direction or just in the direction in which we also already have been applying the force we are really interested in the positive buckling multipliers the second one is also pretty interesting one because this is a site buckling mode but this is also one which I'm not really that concerned with because if we add the pieces of wood the bracket would be more or less constrained in this direction and such a failure would not be that realistic the third buckling mode that we see is one that really seems to be realistic because the same type of failure we have also seen in the the tests of Joel the buckling factor that fusion 360 has calculated is around 1.9 that means that we would need to apply almost 600 Newton's that such a failure could happen okay so with the two analysis we have performed we have identified the critical locations of our bracket that we can just improve in and design optimization so at first we have found in the static simulation that the radius just next to the position of the load introduction is quite highly stressed and also well the lower member of this bracket seems to be critical for buckling not only the simulation is telling us but also the already test results that we have just out of curiosity I have also performed the same analysis on Joel's improved bracket and what we can see is that the stresses have been reduced quite a bit and also the buckling happens quite a bit later and this was also validated by the second row of tests that he has performed in his video okay so now I'll go ahead with the design optimization so we have found out that two locations are critical and we want to tackle both of the problems for the optimization I basically wants to just increase the size of the radii at specific locations that the stress concentration is not that big anymore and I want to thicken a little bit the lower member of the bracket that the buckling failure will happen later and should actually be not a problem anymore for the later tests so I started with the initial design of Joel and I basically just selected all of the surfaces I wasn't really happy with after selecting them you can simply hit delete and they will be gone from the model the first thing I'll do is just increase the thickness of the lower member and you can use the press pool tool for that so just select it from the options select the surface you want to modify and drag the arrow a little bit in my case five millimeters that the thickness is now doubled the next thing is a little bit more usability so I'll just shift the location of the holes that we need to well screw the bracket to the wall a little bit in order to create the holes I simply add a construction line on the back of the bracket and then add the three holes with a diameter of four point five millimeters I will be adding the support in the middle for the lower trust back again by just again creating a construction line that goes to the middle of the lower member and then add some additional two lines to just finish the sketch up after extruding the support I continued with the Philips and basically just select all of the sharp corners and add fill it's in a way that it kind of looks nice usually the bigger they are the better because this results in most of the time less stress concentration and there we have it before we printed out I again just want to run the simulations on this model again and check whether it performs better in terms of simulation in comparison to the two other ones the setup is pretty simple now because we already have the previously well setup simulations in the background and we just need to change some of the boundary conditions in this case especially the fixed supports where the bracket is bolted to the wall I started with the buckling analysis and the results nicely showed that our design optimization was really good because the first real significant buckling mode was at a factor of 10 and this is only the site mode that should be constrained when we later use it as a as a filament rack we actually can be quite sure that we shouldn't see any buckling failure during the test of this new bracket also the results of the static stress simulations we're pretty interesting because the maximum stress we see now is around 16 mega Pascal's which is almost by a factor of 4 reduced to the first design of Joel so I'm quite certain that this design should perform quite better and should be able to bear quite a bit more amount of load before it fails and all of that is add pretty much the same way it in the end next I printed all three designs on my Cyr tan with the Titan arrow extruder and 0.6 millimeter nozzle in plastic at our check yellow PLA in order to compare them later I used exactly the same settings for all of them with three parameters 0.3 millimeter layer height for top and bottom layers as well as 20% infill Joel's initial break had weighed the least with 59 grams then there was the optimized bracket with 67 grams and the heaviest one was Schultz revised version with 71 grams I built myself the same crude test rig where I screwed the brackets to a piece of 2x4 loaded it with a ratcheting strap and measured the load with a simple heavy duty scale the initial bracket nicely failed u2 buckling of the lower frame member just as seen in the simulation at 45 kilograms of load Joel's revised bracket failed at 69 kilograms of load due to side buckling and shattered quite badly my optimized bracket performed the best but failed at 74 kilograms at a location I've actually little omitted due to the boundary conditions I've chosen and this is probably a really important lesson when working with finite element analysis if your assumptions are bad in this case boundary conditions your results will also be worthless always make sure that you check the influence of different boundary simplifications for their effect and always really always validate and correlate your results with tests so in this case we running the simulation without the frictionless support nicely correlates with the test results and highlights the problematic read and gives us the opportunity to improve the design even more in my case I moved the middle hall a little to the top and smooth out the radius even more which results in way less stress in the simulation I haven't tested this design yet but I'm quite sure that this would again outperform the others all right I hope I was able to give you an insight into some simulation options in fusion 360 and that even though it's simple to create pretty pictures requires quite some experience if you liked the video please click the thumbs up button and subscribe to the channel if you want to support the making of these videos consider becoming a patreon or use the Amazon affiliate links down in the description of videos in and I hope to see you again next time [Music]

Channel: CNC Kitchen

Views: 266,748

Keywords: fusion 360, autodesk fusion 360, 3d printing, fem, fea, simulation, stress, strain, joel, 3d printing nerd, buckling, failure, topology optimization, finite element simulation, engineering, cr-10, pla, cnc kitchen, stress simulation, strength of 3d prints, filament holder, filament rack, tutorial, guide, best practice, fusion 360 tutorial, stefan hermann, joel telling, titan aero, study, static stress, structural buckling, buckling mode, buckling factor

Id: LBZAH5RMR2c

Length: 21min 38sec (1298 seconds)

Published: Sun Sep 16 2018