Understanding Force/Stress Outputs in CLT Panel

[Anonymous]

This thread is a follow-up on a previous thread, with a slightly new path (see: http://forums.autodesk.com/t5/robot-structural-analysis-forum/modeling-clt/td-p/6561783)

I have some more verification work to do with modelling CLT panels this way, but one challenge I'm having off the bat is understanding the outputs. Admittedly, I'm not verifying with the simplest case, but I'm in a bit of a crunch to apply this stuff to a real problem.

I've included a quick hand sketch and a screenshot of one of the maps on a panel that I'm modeling. The panel is handing from two bars, with a UDL (dead load) applied at the bottom. I've verified that the reactions in the bars are as expected! that's great news... But I want to understand the stress demands in the panel now. In the attached drawing, I indicate what I believe to be important, including the axes, and I'm hoping to get some insight as to how to view these outputs.

One small additional quirk is that the panel thickness had to be fudged (as per the thread linked at the beginning -- Arthur.Kosakowski will understand). Since it's fudged, I don't think we can rely on the stress maps. We'll have to use only the force outputs and calc our own stresses. This is just my thoughts for now.

I am looking at some other threads, and this RSA Vault link I found in the thread, but I must say that the RSA documentation for the Maps dialogue is not the most helpful, especially for users coming from other software. I'm hoping this thread will give me the insight I need to pass along to my colleagues.

[Artur.Kosakowski]

I'm not sure if I understand you correctly. If we assume that a CLT panel is build from numbers of layers of 'timbers' with different characteristic the stress distribution in each of such panels will be different depending on these layers (each of layer having different stress rather than linearly varying as for a solid material). If so, checking the stress map directly may not have any sense and this should be done outside the program based on in-plane forces and out of the plane bending instead.

[Anonymous]

This is a fair point. I know what you mean because I've been exploring the outputs from Dlubal's RFEM software and RF-Laminate add-in. Dlubal's RF-Lamiante add-in takes the panel stress outputs from RFEM and breaks it up into the stresses distributed throughout the panel layers. But then what does it mean for using Robot for analysing CLT at all? What can I analyse and take away from a CLT panel modelled in Robot?

Edit: I get the sense we are on the same page, regarding the importance of force outputs rather than stress outputs. Perhaps what you are saying is that I shouldn't be focusing on stress maps, but force maps, and then I need to do my own analysis outside of Robot?

[Artur.Kosakowski]

Edit: I get the sense we are on the same page, regarding the importance of force outputs rather than stress outputs. Perhaps what you are saying is that I shouldn't be focusing on stress maps, but force maps, and then I need to do my own analysis outside of Robot?

Yes, this is what I meant.

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

---

[Tuctas]

Using CLT panels in Robot!? If you find the way keep us informed...

[Anonymous]

---

@Tuctas wrote:

Using CLT panels in Robot!? If you find the way keep us informed...

---

It won't be an automated solution -- or even the best software solution -- but as per the comments above, it can be done such that you take the force outputs and compare to the max allowable forces.

If you need CLT modeled and analyzed by one piece of software, Dlubal's RFEM and RF-Laminate are probably your best bet (as per my recent research).

[Anonymous]

---

@Artur.Kosakowski wrote:

Edit: I get the sense we are on the same page, regarding the importance of force outputs rather than stress outputs. Perhaps what you are saying is that I shouldn't be focusing on stress maps, but force maps, and then I need to do my own analysis outside of Robot?

Yes, this is what I meant

---

Okay, so then I'll retry my original question after I get to the next stage of some modeling of this panel, where I will have more complex force maps, and will be trying to interpret the results. That's where I am going to have the most trouble taking the Robot results to get useful results, just because I'm not overly confident with these Maps dialogues, and the documentation hasn't been extremely helpful (basic examples would go so far to help!)

I'll have some diagrams soon.

[Artur.Kosakowski]

1. Set the projection of the results as parallel to the global Z axis (Direction Z instead of automatic)

2. Nxx map will correspond to 'vertical' in-plane force distribution (as for a vertical column)

3. Nyy will correspond to the horizontal in-plane forces

4. Mxx will be bending as for a vertical column

5. Myy will be bending in the perpendicular to Mxx direction

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Anonymous]

That's actually very helpful! Thanks. I was misinterpreting the graphics in the documentation, so that's what was holding me back a bit.

Now that I'm dealing with just this panel, I guess my only concerns are the membrane forces (N). On the simpler side of things, my largest tensile forces are in the Nxx plane, and I've analyzed those as per the attached document. However, the largest compression force are coming out when I select Nxy (the output is also included in the attached PDF). I don't quite understand that particular output or how I'll analyze it.

[Artur.Kosakowski]

Nxy is the in-plane shear.

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Anonymous]

Okay, then is it as simple as the outlined calc in the attached pdf? I have a feeling it's trickier than this...

[Artur.Kosakowski]

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Anonymous]

Thanks Artur,

I did assume a parabolic distribution of shear stress. I realize this requires a constant shear force over that cross section. I appreciate that this may be a small stretch for this particular case, so I will be careful not to make a decision where the demand is too close to capacity. I'm not sure how I would go about evaluating the reality of this assumption though... I think I would need to take a cut through that plane, and examine the shear envelope.

I'll follow up if another question comes up.

Cheers

[Anonymous]

Back at it this week.

I am evaluating the bending stresses now, and based on your previous responses, I don't see a direct way to evaluate bending moments. I think the way to do it is to look at the Nxx and Nyy outputs, through which I can evaluate the resulting tension and compression forces that come from the bending action. 

I've attached the Nxx and Nyy outputs (Local X-axis is vertical with respect to the panel). As noted in the figure, the Nxx shouldn't have significant forces outside the bar connection zones. The Nyy result should have relatively large tension and compression forces, representing the tension and compression due to the bending. The active deformation setting is turned on for visualizing that bending shape as well. I've circled a beam at the bottom too, which I expected to see larger forces on... Maybe I'm misjudging though, because the forces are relatively much higher at the bar connection zone.

[Anonymous]

I've been continuing on this model for a little bit, and getting into the Panel Cut options to try and determine the peak stresses in the panel (especially due to bending now --- I think I'm satisfied with the shear stress calc).

So in addition to my post from earlier today, I have attached a hand drawing of the Nxx and Nyy results from the area next to the connection, and I am trying to understand these outputs; but either I'm missing something about the parameters, or there's an inconsistency in the local system in my Panel Cut and my Maps (although I did make sure to define both as having a local x axis in the same direction as the Global Z direction, as per your previous suggestions).

Anyways, please see attached. I've focused on the connection region. I'm really not grasping Nxx and Nyy in the maps versus Nxx and Nyy in the panel cuts. (and feel free to see the attached RTD file as well).

***Edit: I think I need to restate this for clarity. In summary, I want to know what tensile/compressive stress to design for (tensile likely to govern while this panel is hanging). I think Panel Cuts are the best tool to determine this design force; so I need to know how to understand the outputs from Panel Cuts in order to evaluate that.

[Artur.Kosakowski]

My understanding is that maps (UX blocked in the planar supports) and cuts were already solved during your meeting with Pawel earlier this week.