Help with coordinate sytem

[EdClymer]

Greetings


I have used since 2000; an excellent FEA package called AutoFEA (by JL Analyzer). (It has earned much more than it cost).

When I draw a model, I use standard Cartesian – x, y & z.

When (after processing) I look for bending stress (between two nodes) I am offered ‘S’ and ‘T’ directions (with ‘R’ being along the beam centroid).

The manual says, “beam stress is always in the beam co-ordinate system - that is axial R, and S and T in the beam section”.

The programme was written in California.

Can anyone explain this beam section co-ordinate system
(and interpret to Cartesian).

Regards

Ed





Ed Clymer
Resinfab & Associates
England

 

[NRP99]

Check whether you can create user defined system and use that to post process the results.

You can contact customer support instead of trying to guess yourself//us to better understand this. Genaeally all softwares provide stress in Cartesian or User defined co-ordinate stress.

 

[rb1957]

if the program is so "excellent" then this should be explained better.

a beam has it's own axes. typically a beam is a two noded element, and one axis is between the two nodes, axial but which way (sometimes it's N1 to N2 sometimes the lower node number to the higher. this is typically the "X" axis, in this case the "R" axis. the "Y" axis is orientated towards the 3rd point (that you've specified somewhere, typically on the property card).

Maybe consult an FEA text, this'll be explained there too.

Failing that play with it ... with some known geometry and see how it treats that.

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[ThomasH]

I Googled "AutoFEA" but that did not give any useful info. The software has been mentioned in the forum before but the website that was mentioned then seems to be gone.

Regarding the coordinat system. I would start with a very simple cross section and three load cases, one with axial load and two with moments (bi axial). That should be fairly east to compare to a hand calculation.

But I can't help wondering, what type of analysis have you used the software for? This is not the type of question I would expect after more than 20 years usage.

 

[EdClymer]

Greetings

My thanks for kind replies.

NRP 99. I purchased a CD containing the programme in 2000: by 2002 the supplier had gone bust. They only communicated
in Spanish. It is a very good FEA, it always passes validation. There is no way to change orientation system. Geometry is in
Cartesian. Beam section stress is not in 3D Cartesian.

Rb1957. You are correct; beam is a two noded element. The manual specifies ‘R‘ as length (drawn in Cartesian as ‘z’). After processing I am offered ‘S’ and ‘T’ directions of bending stress – without indication, of which is which (in Cartesian).

I have ‘played with it’a rectangular section)and concluded (a long time ago) that ‘T‘ was for top load, vertical, (ie y direction): And ‘S‘ was for side load, horizontal (ie x direction).

The manual says; “beam stress is always in the beam co-ordinate system - that is axial R, and S and T in the beam section”.

I just wanted to know; what system uses R, S & T and how to convert to Cartesian.

“Excellent” means it gives correct answers: Manual was written in Spanish and then (badly) translated.

Regards

Ed




Ed Clymer
Resinfab & Associates
England

 

[ThomasH]

I asked ChatGPT.

There is a system used in Spain called R, T, S coordinate system. R stands for radial, T for tangential and S for normal. It is often used for curved or circular elements. It is not a Cartesian system in the traditional sense.

 

[IDS]

ThomasH - it doesn't sound like it is being used like that in this case, since we are talking about the local member axes, which is a rectangular system.

I have a blog post about converting from local to global axes (and vice versa) at:

https://newtonexcelbach.com/2014/01/08/converting-from-global-to-local-coordinates-and-vice-versa/

Please ask if any questions.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[rb1957]

R, S, T are just 3 letters, like L, M, N. They may mean something in Spanish, but then I don't know Spanish.

You say you've done some testing so you've figured out ... R is axial, S is one transverse direction, T the other. These are cartesian co-ordinates (in that they obey the RH rule (index is R, middle is S (or T), thumb is T (or S).

I don't understand the question ?

If you don't define a 3rd point (to set up these axes, I wonder how the program does it ? I would be very careful with the testing to decide. Maybe S is always strong axis ? Does it assume the beams are rods (ie circular), so directions don't really matter ??

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[ThomasH]

ThomasH - it doesn't sound like it is being used like that in this case, since we are talking about the local member axes, which is a rectangular system.

I agree, it does not sound like it is being used like that.

What I did was to write a short description based on the info from the OP and "ask" ChatGPT. And I also mentioned that it could be of Spanish origin since the manual was in Spanish. And ChatGPT returned the description that there is a Spanish system with the same terminology used for curved or circular beams. It may be a coincidence that they use R,T, S for the axis. But I find it surprising that this question comes from the OP after such long use.

 

[NRP99]

Try solving your beam problem in general purpose FEA software like Ansys/abaqus and compare the results with "AutoFEA". See what results you are getting at each of the axis direction.

 

[rb1957]

Compare FEA with a handcalc ?

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[EdClymer]

Greetings

My thanks to all who contributed to my request for help, with special thanks to ThomasH who confirmed (aided by AI) that such a direction system of R, S & T existed and it obeys the Cartesian RH rule.

I just could not envisage an alternative to Cartesian (or why);particularly as initial
geometry is in Cartesian (x, y, z).

The FEA system is used for the design of small and large plastics structures, or components
of mostly plates and shells.

I did a ‘hand calculation’ of a simple cantilever beam (via MathCAD) so I know what the correct deflection and stresses are.

I have always taken ‘S’ to mean when force applied from side: and ‘T’ when force applied from above when working with sections. The FEA returns a maximum bending stress of 8.32e001 MPa and zero; when MathCAD says 83.3 MPa (in the correct logical direction) and zero in the other.

So I think QED.

Well done and thanks to all (and apology for late reply).

Regards

Dr Ed


Ed Clymer
Resinfab & Associates
England