Guardrail Glass Engineering Software

[lanejoe10]

What software are people using to analyze glass used in a guardrail system?

 

[KootK]

1) Any FEM package with plate capability (SFrame, SAP, RISA, etc).
2) MEPLA. Powerful but expensive and user hostile.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[lanejoe10]

KootK-
Thanks. We have been using Risa as well which does great for monolithic pieces of glass but we are having a hard time trying to figure out how to have it accurately analyze laminated glass.

Thanks for the recommendation on MEPLA, we will look into that.

 

[KootK]

For laminated glass, there are procedures available in ASTM 1300 for determining an equivalent glass thickness. That equivalent thickness can then be used in Risa etc.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[lanejoe10]

Ahh yes Appendix X9. Hadn't got that far through ASTM E1300 yet.

Thanks KootK!

 

[KootK]

Happy to help lanejoe.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[NorthCivil]

KootK,

Afraid I disagree here.

1) Any FEM package with plate capability (SFrame, SAP, RISA, etc).

Plates that experience deflection greater than 1/2 of the plate thickness are more accurately modelled as a shell to take into account membrane stresses.

2) MEPLA. Powerful but expensive and user hostile.

I think I remember discussing Mepla with you fairly recently... Its actually a very user friendly program, but IMO, returns poor results. I've done a bit of research since into why. SJ mepla uses Mindlin plate theory, which is actually more applicable to plates of a span less than 10% of the thickness.

I have since been on a bit of a foray into plate/shell theory, which makes for some heavy weekend reading...

And lanejoe:

please, for the love of god, don't give into the pressure of the glazing contractor and sign off on 12mm tempered guardrail... it seems to be a bit of an epidemic around here, and is definitely unsafe. 8on8 temp lam works a lot of the time, 19mm temp with a structural top cap, or 10mm on 10mm temp lam can take just about anything in a commercial application. These are good rules of thumb to get you started.

 

[KootK]

Afraid I disagree here... Plates that experience deflection greater than 1/2 of the plate thickness are more accurately modelled as a shell to take into account membrane stresses.

Semantics. All of the listed packages will do shell when deflections warrant it.

I think I remember discussing Mepla with you fairly recently... Its actually a very user friendly program, but IMO, returns poor results.

What?!? Okay, here we massively disagree. I've used every FEM package under the sun. Mepla's user unfriendliness stems from:

1) The only units available newtons and millimeters.

2) Chunks of the interface, the warnings, and the manual are basically in German.

3) The GUI appears to be Windows 95.

4) You draw stuff by entering coordinates in a strict counter clockwise fashion. Seriously. In 2016. Not even DXF import.

5) The only way to get precise output at a particular location is to specify that location in coordinates before running the analysis.

6) The only way to get to a particular load case and output form is to cycle through all of the load cases and output forms.

7) The only way to handle files is to dump and load everything as goofy zip packages.

8) If an error occurs in your model, you get to find it yourself based on coordinates provided.





I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[NorthCivil]

Semantics. All of the listed packages will do shell when deflections warrant it.

With glass, deflections always warrant it. In fact, I can't think of a single time I have used FEM, critical to the design of a glass lite, that didnt return a deflection equal to or greater than the thickness of the lite. Membrane stresses are very important (and it goes without saying, modelling should always be non-linear).

Regarding SJ mepla and its user interface. Difference of opinion - I don't think we're going to come together on that one. I think it is great in its old DOS-y feel. Very easy to edit a model. Very quick to set up parameters. Not surprising, as its inputs revolve fully around practical glass applications.

But that aside. The main flaw IMO with SJ mepla is its use of mindlin plate theory, a theory not well suited for glass analysis. This is the reason SJ Mepla became dusty-shelf-ware at my previous firm, a firm that practiced exclusively in structural glass. The results it returned did not correlate with what we had seen empirically, or with better performing models run in other FEM packages.

 

[klaus.]

Mepla is very good and fast if you are doing glass cals every day
Results are accurate enough ... for practical design ...
For academic cals a full FE program (like Ansys) might be better but also much more expensive
Yes it has certain things which one has to get used to....like the N and mm

RFEM now also has a addon module for glass cals....including the PVB stiffness and gas Volume

But yes...most other FE software is good to ....most important is, taht the user knows what he (or she) is doing

my first glass calc I used formulas for concrete slab design.... glass is still up and not broken


best regards
Klaus

 

[Yao1989]

I do not mean to necro old threads, but be careful using RFEM's RF-Glass module if you are running a "global" model. The global model assumes full composite action by default, and cannot calculate partial composite action.