Finite element analysis results interpretation

In summary: So I do not understand the situation.In summary, the conversation discusses the results of a finite element analysis of an automotive component and the concerns regarding the stress and displacement values. The analysis shows that the stress is beyond the material's yield strength and there is potential for permanent deformation or failure. Different viewpoints are shared on how to interpret the results and potential solutions are suggested, such as adding additional stiffeners or changing the shape of the existing ones. The need for further analysis and considerations for cost are also mentioned.
  • #1
gonzalo75
4
0
Hello,

I have been following this forum for some time now, but this is the first time I participate. I would appreciate if you could give me a hand in understanding the results of a finite element analysis of an automotive component we have designed.

The analysis results indicate that the stress caused by the load applied is way beyond the material yield strength and even the ultimate tensile strength. The values are 258MPa, 90MPa and 150MPa respectively. This is shown in the first image attached.
Considering this I would assume that not only the part is subjected to permanent deformation but it may also break.

However when I look at the displacement results I notice that the part only buckles 0.1mm. Second image attached. This deformation represents only 0.5% of material elongation. Knowing that the aluminium alloy we are planning to use breaks @2% elongation, does the analysis really show that the component will fail or are we just looking at a hot spot that will not cause any problems?
The part will not be subjected to any cyclic loading.

I would appreciate some feedback. Thanks for reading.

Regards
gonzalo
 

Attachments

  • stress.jpg
    stress.jpg
    71.6 KB · Views: 1,555
  • displacement.jpg
    displacement.jpg
    47.9 KB · Views: 1,510
Engineering news on Phys.org
  • #2
gonzalo75: The strain corresponding to the yield stress is approximately 0.13 %. The analysis indicates that the general stress on the part currently exceeds the material yield strength. It does not appear to be just an insignificant hot spot.

It is better for the general stress on the part to not exceed, nor get near, the yield strength. The analysis is currently telling you that you need to increase the size and/or thickness of some stiffeners. The analysis is also telling you to increase the fillet radii in a few spots.
 
Last edited:
  • #3
Hi gonzalo, welcome to the board. nvn always has terrific posts and in general I agree with the assessement, but I'd like add a few thoughts.

If this is a cyclic stress, then you should determine the maximum and minimum and look at fatigue as well. Typically, cyclic stresses above yield lead to fairly rapid deterioration and failure of the part. That isn't necessarily true of static stresses though and the part may be ok as is. Just because stresses are above yield in some localized area as you show, the part isn't necessarily going to fail. What can happen is there will be some localized yielding which results in increased stresses for material farther away from the peak stress areas. So when you did your analysis, you assumed the stress/strain relationship was linear and therefore the resulting output does not predict yielding and doesn't give you a representation of what might happen farther away from the peak stresses for the actual part.

I'm not an expert in FEA but I do work with our FEA engineers regularly and I know we have a feature which allows you to limit stresses in the elements to yield, at which point the model basically assumes the modulus goes to zero and the material just stretches at that given stress level. The model then works out how surrounding elements will react to this stretching. The stress in those surrounding elements must increase, and if they increase to the yield point, they continue to stretch. So you could look at it that way and see what the total deformation and final stress state looks like and see if that's acceptable.

Note that yielding in a part like this isn't unusual at all. Parts that are manufactured by forming undergo this same yielding process locally. Things like structural beams, sheet metal and piping are commonly shaped by permenantly deforming the part through localized yielding. The resulting stresses are generally not removed by heat treatment and remain in the part for life.

This same philosohy is accepted by ASME for example, when calculating bending stresses on piping systems due to thermal contraction or expansion. ASME piping codes allow yielding in those systems during operation, knowing that the deformation and thus the stresses can be limited to a single event. That is, the stresses can be made such that they only exceed yield once during operation and subsequent thermal cycling results only in the part going from one stress level to another without yielding a second time.

So basically, you should look at cyclic stresses if they exist and determine if this is a fatigue issue, in which case I would suggest changing the design to reduce stress to well below the yield point. If not, you may consider changing your model to allow for yielding and find out what happens to the final part, then determine if the resulting deformation is acceptable or not.
 
  • #4
Hi,

Thanks a lot for your feedback. It really helps.

The load applied will be static. It corresponds to the load applied by a bolt, which once tightened it won´t be removed anymore.

Based on your feedback I will add an additional rib just like the two taking the stress right now.

Perhaps the current design is ok and just some yielding would make the excess of stress go away, but since the component does not exist yet it takes no money or effort to introduce the change at this stage.

Currently we outsource these analysis. These don´t come cheap, so there is a very limited number of iterations I can afford.

Cheers
gonzalo
 
  • #5
gonzalo75: Another possible issue is the inefficient shape of the stiffener. The current concave profile of the overstressed stiffener causes material to be missing, where material is needed. If you instead use a straight (not concave) stiffener profile, with only a slight (not rapid) depth taper, then the stiffener might be much more efficient.

What aluminum alloy (and temper) is this? Unless I am misinterpreting, a 2 % rupture strain sounds brittle, in which case stresses greatly exceeding the yield strength might not redistribute well.

You say it would take no money to make a change now, but you said the analysis is very expensive. Therefore, unless I am misinterpreting, it sounds like your posted stress analysis is not one of the expensive, outsourced, final analyses, right?
 
Last edited:
  • #6
Hello nvn,

The stiffener was originally much more substantial, but it resulted in a stress concentration right at the corner. This was shown in the first analysis we did, which I have attached. Peak stress was then even higher than now. That was when we decided to make the stiffener more progressive to distribute the stress. The resulting stress certainly dropped. But it is still beyond yield.

Part of the problem is that the axial load applied by a bolt is incredibly high. For example, in this case we are using an M10 x 1 (fine thread) bolt. A torque of 10Nm, which is not that high, results in 4800N of axial load! Initially I did a calculation using T = 0.2 D F (D being the bolt diameter. Units in inch and feet) and thought that the result could not be right. Therefore I conducted a test in the lab with a load cell; The theoretical result was only 200N away from the real value.

The aluminium alloy we are using is AlSi12Cu, which is quite common for this kind of application. I also find that a 2% rupture strain is quite low, even more after having tortured a similar part injected with exactly the same material. I have attached a photo in which you may see how much the part buckles without failure. After that the part never recovered, but I was not expecting that much bending without failure.
Having said that, I need to stick to the material properties provided by the manufacturer.

The stress analysis I posted in the first message is indeed the second analysis we outsourced. The first analysis is the one I am attaching now. They are expensive unfortunately, so optimization through trial and error is not an option.
It takes no money to introduce any changes now because the injection mould has not been released yet, so any changes we do now are simply CAD data changes, which we do in house and are “free”. It is running unlimited FEA with the changes that we cannot afford.

Regards
gonzalo
 

Attachments

  • initial_design.jpg
    initial_design.jpg
    68.5 KB · Views: 1,017
  • test.JPG
    test.JPG
    41.1 KB · Views: 937
  • #7
That's amazing! I was envisioning this as being some fairly large casting, the size of a toaster or something like that. Then you show a small 10 mm bolt in the center where the rectangular hole is.

I wonder why all the detailed gussets and things are in that part. Are they there for a reason? What would stop you from making this out of essentially a flat piece of material with a hole drilled in the center, perhaps with a countersink to keep the head of the socket head cap screw from protruding too far?
 
  • #8
Yes, the part is not very large at all. It is roughly 90mm x70mm.
It has all those ribs and holes because it interfaces with several other components. It is actually the lifter of a window regulator. It basically clamps the glass and moves it up and down along a guide.
I wish I could make it simpler but the interface is complex.

Regards
gonzalo
 
  • #9
The original design (from post #6) has a very wide scale, so it's difficult to see how far the peak stress actually extends from the corner, though it clearly won't extend very far because of the geometry. Also, that stress is a compressive stress which is a good thing. I'd love to see a printout of the original design using the same scale as the new design. I suspect the amount of material that is actually above yield stress is smaller with the old design when compared to the new design. Being as it's in compression and it's static, and assuming the amount of material above yield in the old design is less than in the new one, I'd think the old design might be better quite honestly. But without knowing how much material is truly above yield, I don't think you can properly compare the two designs.

To reduce stress, I don't think you'll be able to simply change the design of the ribs that are exceeding yield without filling in the hollowed out areas. A higher moment of inertia would help dramatically though, so making the area where the screw goes through thicker (and the ribs taller) would make a dramatic difference. I suspect if you doubled the thickness of the material where the screw goes through, your stresses would drop to well below yield at all locations. The problem may be that the head of the screw then sticks up too far. The solution to that is to get a low head, flat head or button head socket cap screw or similar low profile head screw. The more material you can add under the head to make your part thicker, the lower the stress will be. And the relationship between thickness at that location and stress won't be linear - doubling the height should reduce the stress by approximately 75%.
 
  • #10
gonzalo75: I, too, currently think I might prefer the initial design, rather than the second design. In addition to the suggestions by Q_Goest, in the original design, you might want to make the following changes.

(1) Make the fillet radius larger at the stress concentration (where the red hot spot is currently shown in post 6).

(2) Add one stiffener, like the two existing stiffeners, which will give you three stiffeners, instead of two.

(3) Make all three stiffeners substantially thicker than the current stiffener thickness. There is virtually no constraint on stiffener thickness; therefore, you should make all three stiffeners much thicker.​

It would be interesting if you show us your next iteration of stress results.
 

1. What is finite element analysis (FEA)?

Finite element analysis (FEA) is a computer-based numerical method used to solve complex engineering problems by dividing them into smaller, simpler parts called finite elements. It is widely used in various fields of engineering, such as structural analysis, heat transfer, fluid dynamics, and electromagnetics.

2. How are FEA results interpreted?

FEA results are typically presented in the form of visual representations, such as contour plots, stress plots, and displacement plots. These plots show the distribution and magnitude of various parameters, such as stresses, strains, and deformations, across the analyzed structure.

3. What is the significance of FEA results interpretation?

Interpreting FEA results is crucial for understanding the behavior of a structure under different loading conditions. It helps engineers identify potential problem areas, optimize designs, and make informed decisions about the structural integrity and performance of a design.

4. What are the limitations of FEA results interpretation?

FEA results are only as accurate as the input data and assumptions used in the analysis. Therefore, it is essential to carefully validate the model and ensure that it represents the real-world conditions accurately. FEA results may also be affected by various numerical errors, such as element distortion, mesh density, and convergence issues.

5. How can one improve the accuracy of FEA results interpretation?

To improve the accuracy of FEA results, engineers can use advanced techniques, such as mesh refinement, convergence studies, and sensitivity analysis. Additionally, it is essential to have a thorough understanding of the underlying theory and assumptions of FEA and carefully validate the model against physical tests or analytical solutions.

Similar threads

  • Mechanical Engineering
Replies
2
Views
839
  • Mechanical Engineering
Replies
3
Views
2K
Replies
4
Views
706
  • Mechanical Engineering
Replies
1
Views
1K
  • Mechanical Engineering
Replies
9
Views
1K
  • Mechanical Engineering
Replies
2
Views
1K
  • STEM Academic Advising
Replies
5
Views
1K
  • Mechanical Engineering
Replies
2
Views
3K
  • Mechanical Engineering
Replies
3
Views
9K
Replies
3
Views
2K
Back
Top