First of all, you have to understand for polymer, the yield stress might in a wide range which depends on manufacturer and batch, I will recommend you to get the relative information from www.matweb.com , and you have to choose from one manufacturer and the lower value of yield stress should be used.
After that, you have to give a large allowance of FOS due to the property of Polymers:
- Plastics will deform permanently under load even the load is relatively low.
- When subjected to static low stress / strain a ductile / brittle transition will occur at some point in time resulting in brittle failure
- Cyclic stressing will result in a ductile / brittle transition resulting in brittle failure at low stress level
- Premature initiation of cracking and embitterment of a plastic can occur due to the simultaneous action of stress and strain and contact with specific chemical environments (liquid or vapor)
Polymer failure modes
- Deformation and distortion due to creep & stress relaxation,
- Brittle Fracture due to Creep rupture (static fatigue),
- Notched creep rupture, Fatigue (slow crack growth from cyclic loading), High energy impact
- Wear & abrasion,
- Thermal fatigue
- Degradation – thermo-oxidation
- Dimensional instability
- Additive extraction
- Solvation, Swelling, dimensional instability and additive extraction
- Acid induced stress corrosion cracking (SCC)
- Hydrolysis (water, acid or alkali)
- Environmental stress cracking (ESC)
We are making many simulations for plastic par ( PA6, PP, PBT or PA66 for example ) I have already posted a picture showing the limit to considere depending on the art of stress that you have. A plastic is up to 0,2 % strain elastic and then start the plastic area. The problem ist how long at which temperature !. The picture I have posted give you some indication for a part design to withstand 10 to 15 years.
We are working for an automotiv supplier and already test and confirm those limit many time
Can you explain in general terms how you determine the allowable stresses for the plastic. The graphs don't help me much.
To find a yield stress I understand you use a 0.2% strain offset. Knowing the yield stress, the allowable stress could be 2/3 x Yield Stress ie. FOS = 3/2 or 1.5.
What kind of safety factors do you use and how do you determine these?
so you do not have any offset, the 0,2% is just more or less the theorie.
to explain the graph let s take an example.
1) You want to design a plastic housing ( PA66GF30 for example ). The housing will be under pressur coming from a spring and the temperatur is from -40°c up to 120°c.
In that case we are taking the strain/stress curve from the PA66GF30 @ 120°c and looking for a maximal stress around 0,5 % ( like the graph say ).
2) If there are exeptional force comming from an other source, we will do the simulation with a maximum allowed stress at 1% Strain
3) if the Housing have some cliping function, we will simulate the bending of the clip with a maximal stress of 0,5 x Max Strain @ room temperature.
The safety coeficient or FOS is not really possible to calculate because it will not only depending on the fiber orientation but from the" cold joint" as well !
we have a mold simulation software integrated in Solidworks ( Simpoeworks ) and doing a mold simulation before stress simulation to see a little bit what happen.
So the main concern is the Stain level but not the stress level?
And the Stain level is the main contribution factor of creep?
The strain give me the allowed stress.
By the way, I have to add something, just looking the Von mises is not enough ! you have to look the principal stress as well, at least the first !.
Let me give you an exemple, if a volume of the part you are simulating is growing in the way that the shap stay exactly the same but is bigger ( like a zomm effect , a ball stay a ball but bigger ) the Von mises will be zero !! ( yes look the formel to understand when all stress a equal ).
but the first Pricipal stress will gibe you the value.
to understand the what is the principal stress :
For a 3D element (a cube) there will be 3 normal stresses and 3 shear stresses. They are often given in a global coordinat system.
Now imagine that you create a local coordinat system for each element. And imagine that you rotate the coordinate system so that there is no shear stresses on the cubes surfaces. The only stresses present for the rotated cube are "pure" normal stresses. Those are the principal stresses.
So my advice :
in Solidworks Simulation, always compute the Von mises and first Principale stress
Okay thanks, that helps to explain what the graphs are about. Seems to me that you are setting your limiting criteria as strain like Shaodin says. But I can see that you can get an allowable stress based on the strain. You talk about fibre orientation in a previous post. I understand that strain criteria is usually used when working with fibre reinforced plastics. I'm not sure if that is a normal apprach when talking about a polymer.
To be honest with you I do not undertsand your question, because Polymer refer to a very large class of synthetic materials. So the reinforced "plastic" are included insid the Polymer Name. What do you mean exactly when you said Polymer ? you mean without Fiber ?.
I'm referring to platics like HDPE and expanded polystyrene with out fibre reinforcing.
I see, I have to say that for those polymer I have no experiences, but one is sure, the model you will have to take in Solidworks simulation as to be controlled. As you know, not all plastic are using the same Model, for example I m using the Hyperelastic sometimes as well which is for TPE.
Not sure about you comments regarding checking principal stresses. Von Mises stresses are all about the stresses due to the distortion of an element. If the body doesn't undergo distortion then the stresses are not a problem. In the example you gave, the cube with the same stress on each face undergoes no distortion. Von Mises theory suggests that in this scenario the principal stresses which are all the same will not be a problem. The theory suggests that "a body that is stressed in this way can withstand enormous hydrstatic pressures without damage". [refer R.C. Juvinall "Fundamentals of Machine Component Design" 4th edition pg 246]
It s really interessting what you said about the hydrostatic pressure, I will try to see if I can read it.
but from the experience that I have und regarding some discussion I had with Proffesor in some seminar about no linear simulation the Problem of the Von mises formel is really there.
If you take a cube and pull all the 6 surface with the same force the Von mises result will be 0 but the material will be under stress, the question if the material will withstand or not are still open, but you agree on the fact that the stress is there, so the question is which stress ? and the answer is the principal stresses.
the other problem is when the stresse is not the same but very close to another, in that case you think that everythink is acceptable and in the reality the part will break. The principal stress as to be checked every time that you are close to the break limit.
thanks for your coment
Interesting. I can see your point. Have to say I'm not sure on this one.
Thanks for your responses Shaodun. My topic heading may be misleading. You have provided a lot of infromation regarding the modes of failure. I'm trying to focus on criteria to set when doing FEA on a plastic model. The main criteria would limits on stress and strain. I've provided an example of a way to specify a limit on stress ie a qurter of the yield stress. Assume for the moment that the various properties of a plastic are known ie stress strain curves, Young's Modulus, Shear Modulus and Poisons ratio. What do you set the stress limit at to avoid creep? Are you aware of any strain limiting criteria for polymers. I've listed the polymers that are being used in the original post.
Frankly speaking , I do not have much exprience in Polymer FEA, but my major in college is Polymer processing, so base on my experience, I will prefer to have 2~4 times of FOS for stress and strain. If your part is made by extrusion and machining, you can use lower FOS, if your part is made by injection molding, the FOS should be higher.
Polymer is complicated material, I will not prefer use FEA to verify the design for long term usage.