Photoelastic Stress Analysis 2

[MickFlynn]

I would like to do a dynamic stress analysis of a dovetail joint assembly. I am curious about photoelastic stress analysis and how it works exactly. The whole dovetail assembly is about the size of a dice and is made from Inconel 718 and a stainless steel. Is there a 'simple' way to do photoelastic stress analysis. This is just a once-off analysis.

Regards

MF

 

[JWB46]

For a quick overview of the subject you might have a look at Theory of Elasticity by Timoshenko. This is esily available in any library.

There are a number of books deboted to photoelasticity, but will be more difficult to find because hardly anyone does it anymore.

You might find a laboratory somewhere able to help. I don't think you'll get anything you couldn't more easily get from FE analysis, but I would be interested in knowing how you get on - for old times sake !.

Good luck.

 

[EnglishMuffin]

I have a book entitled "Photoelasticity for Designers" by R.B. Heywood - as of this posting there is still a copy available here :

http://dogbert.abebooks.com/servlet/BookSearchPL

But for a one-off I can't imagine why you would want to do this. The technique has pretty much been superceded by FEA, and it's getting harder to find the equipment.

 

[ThomasH]

I am by no means an expert in photoelasticity but might be able to giva a short description to help you decide.

The general idea is based on that if you apply load to a piece of for example glass and view it when polarized light is transmitted through it a colour pattern due to the stresses can be seen. Depending on material etc different colours mean different stress levels.
This means that you have to build your assembly in some kind of glass or celluloid material to be able to view it.

Now, back to your problem. If this is a "once-off" analysis I would recommend this:

If you have access to FEM software and experience in its use it is probably a better tool.

If you don't have access and/or experience find somebody who has and ask for help. This is probably the best and quickest way to get reliable results.

If you still want to use photoelasticity just out of interrest or "for fun". You can find some information in Timoshenkos: "Theory of Elasticity". The method have probably been "killed" by FEM but it has a pedagogic value in my opinion. If you have seen the colour pattern in a "slab" you remember it.

I don't know if this helped you. If you decide to "go ahead" if would be fun to hear how it works out, just out of couriosity.

Good luck and Regards

Thomas

 

[MickFlynn]

Thanks all.

It was basically just a curiosity thing. I have done FE Analysis on the assembly but since I am mechanically testing it anyway, I thought it would be interesting to find out the stresses in the material while testing. Also, I thought that a photoelastic coating could be applied to the actual assembly so that I wouldn't have to make a model, from glass say. But from what I understand, there are reservations with this coating method.

Regards

MF

 

[jetmaker]

MickFlynn,

I have worked with photoelastic coatings before, and I must say that they are very useful for getting a full field view of the structure. In the majority of the cases where I have seen it used, it has been on large monolithic structures, and not detailed assemblies as you have described.

There are 2 approaches here you can try.
1) make the parts out of the photoelastic material. Very useful for getting through-thickness stress distributions. However, the difference in material properties compared to your actual part needs to be evaluated if you are looking for quantitative results. If you want qualitative results, this is the way to go.

2) apply a surface coating, with each part having a discrete coating sheet. There is an art to this, and it takes a little practice. Furthermore, if the geometry of your parts is complex, then it becomes a lot harder. This method is very usefule for obtaining full field results, and can be used in conjunction with strain gauges to refine the results.

Too many people use FEA for their definitive results, yet do not adequately verify their models through simple means. Most of us know that the models are capable of rendering garbage if the proper constrains or assumptions are not applied. Therefore, since you are testing anyways, I would advise the use of photoelastic or strain gauge surveys to verify your FEA results. If a few locations match well, then one has more confidence in the FEA.

Regards,

jetmaker