Plaxis 3D plate output unit kN*m/m

RE: Plaxis 3D plate output unit kN*m/m

Nam Nguyen Duc — Wed, 24 May 2023 09:52:00 GMT

Thank you Stefanos

RE: Plaxis 3D plate output unit kN*m/m

Stefanos Papavasileiou — Mon, 22 May 2023 14:58:00 GMT

Dear Nam,

As you can see in PLAXIS Output, the plate object comprises of several parts. PLAXIS computes the deformations per element. If you need to consider one value for the whole footing, there is a decision to be made by you as a modeller.
Here you can see that part of your footing settles more than the edges.

What would then be your single value for a real case as such? As a designer, you should be able to answer that better than any of us.

RE: Plaxis 3D plate output unit kN*m/m

Nam Nguyen Duc — Mon, 17 Apr 2023 19:46:17 GMT

Hi all,

I am using Plaxis 3D to estimate the settlement of a strip footings. I export the table of Uz displacement of a cross section through the middle of the footing to get the settlement. There are three elements: Plate\_1\_1 (El. 3), Element 1-3 (Plate) (El. 3), (Strip\_footing) (El. 3) as shown in the picure below. So which one I should consider to take the settlement? Thank you,

RE: Plaxis 3D plate output unit kN*m/m

Dennis Waterman — Fri, 06 Jan 2023 10:23:40 GMT

Dear Rafi,

No, each node has only an M11. M11min and M11max are the lowest and highest value of M11 in the plot, over all the nodes.

So if you have a model with 5 nodes with values of M11 of 100,120,90,160 and 110 kNm/m, then M11min = 90 and M11max = 160.

With kind regards,

Dennis Waterman

RE: Plaxis 3D plate output unit kN*m/m

Rafi Keshvary — Fri, 06 Jan 2023 10:19:55 GMT

Hi Dennis,

thanks for the reply. i modelled a retaining wall as plate in Plaxis 3D and as i can see the plate consists of several elements, each element had 6 nodes assigned to it. & for each node there are M11, M11 min & M11 max (the same results for M22 and other normal and shear forces). What i have difficulty to understand is how a node can have max and min bending moments?

we issued the results to client and they are asking explanation for this and if they should use M11 max or only M11?

Best Regards,

RE: Plaxis 3D plate output unit kN*m/m

Dennis Waterman — Fri, 06 Jan 2023 09:41:16 GMT

Hi,

M11 is based on the stresses that result directly from the calculation. M11min and M11max are just the minimum and maximum value of M11 in the current plot.

With kind regards,

Dennis Waterman

RE: Plaxis 3D plate output unit kN*m/m

Rafi Keshvary — Fri, 06 Jan 2023 09:05:35 GMT

Hi,

I'm also very interested in the 3rd question that Renato asked about M11 min & max and how M11 is calculated? it doesn't sound like being average of the two. Is it related to thickness of the structure, e.g. showing moment in the inner and upper side of a plate element?

RE: Plaxis 3D plate output unit kN*m/m

Shan Tom Wong — Sat, 02 Jul 2022 02:33:50 GMT

Hi Dennis & Stefanos,

This discussion is helpful. For clarification, is it also correct that to get the axial force and shear force at a specific location in a 3D pipe/plate elements, the same process applies (which is spotting local peak value, generate a curve and integrate over a specific length)?

In this sample section below, the peak minimum is about -230 kN/m. When I integrate the Q23 over the ~0.6 m length (shaded blue), the total Q23 is approximately -70 kN over 0.6 m length. Is my interpretation correct?

RE: Plaxis 3D plate output unit kN*m/m

Stefanos Papavasileiou — Mon, 07 Feb 2022 09:36:44 GMT

Dear Mikael,

That is exactly what our Curves manager do. You can select a post-calc node on a plate and plot the bending moment vs. dynamic time.

Note that the more steps you have saved, the more detailed the curve plot will be. That is a prerequisite for post-calc points as they depend on the Max number of steps stored setting in the Phases window.

Useful read: communities.bentley.com/.../selecting-points-for-curves

RE: Plaxis 3D plate output unit kN*m/m

Mikael Manuit — Sat, 05 Feb 2022 08:10:39 GMT

Hi again!

May I please ask, is it possible to plot the integrated Bending moments at one specific node versus Dynamic Time?

Or do we need to plot the multiple results of the desired specific node and then do averaging?

Thanks!

RE: Plaxis 3D plate output unit kN*m/m

Renato Reis — Tue, 25 Jan 2022 02:17:45 GMT

Hi Stefanos,

Continuing the discussion I would like to ask you:

1) I was checking the Plaxis3D Structural Verification for Plates and Shells (Plates:https://communities.bentley.com/products/geotech-analysis/w/plaxis-soilvision-wiki/45502/bending-of-plates Shells:https://communities.bentley.com/products/geotech-analysis/w/plaxis-soilvision-wiki/45505/performance-of-shell-elements) and noticed that both verifications consider the plate self-weight as zero. Using the same models I considered the self-weight of the plates and the results were not the same. Using the plate self-weight it was necessary to multiply the max moment by the length of the plate to get the correct value, similar to what you have described. Why this difference? I believe is due to the thickness (d) that I am considering in the plate properties that must be in accordance with the desirable load of the plate. But I would like to confirm this with you.

2) I am modelling a vertical tunnel shaft. I am a little bit confused with the discussion about the integration of the results. It is ALWAYS necessary to multiply the moment results by the plate length (perpendicular to the moment)? Or this is applicable only in cases that I want to simplify and turn out the plate into a line?

My model is illustrated below. This is only one panel of the shaft. This panel is 5m wide and 50m deep. The M11 value that PLAXIS is illustrating is the final value or do I need to multiply it by the length of the panel?

3) For the Plates results PLAXIS gives the M11, M11 min and M11 max. How does PLAXIS calculate the M11?

Thanks!

RE: Plaxis 3D plate output unit kN*m/m

Mikael Manuit — Wed, 19 Jan 2022 14:14:16 GMT

Sir Stefanos, thank you so much for the detailed explanation.

Get Outlook for Android

RE: Plaxis 3D plate output unit kN*m/m

Stefanos Papavasileiou — Tue, 18 Jan 2022 16:32:34 GMT

Dear Mikael,

I believe that you have misunderstood my answer and it is important to fully understand what we mean when we integrate the structural forces as this is fundamental when checking the results of your model.
It is very critical that you review your extraction of results with a fellow colleague from the geotechnical or (even better) structure team to help you understand this concept of integrating structural forces of surface (plate) elements.

To start with, in your original question you asked about integrating the structural forces that PLAXIS gives for a surface plate element to consider/simplify it to a line element. The way to do this is by using the current width of your tunnel slice, which you specified is 0.06m.
As I mentioned before, when we do this integration over a slice width, we assume that the results are constant over that width. Then our integration can become a simple multiplication of the result value and the slice width, which in this specific case is 0.06m.

The number "0.06m" is no special number but rather refers to the slice width for that specific integration of forces.

PLAXIS Output always returns the resulting bending moments in the units of kNm/m as it refers to a surface element and not a line.
If one wants to consider simplifying to a line element (from a surface element) an integration of the forces over the width (perpendicular direction to the slice) is required.

In your latest post, you mention that you would change the model and the tunnel slice instead of 0.06m would now be 1.00m. The results you will see in Output will still be in the units of kNm/m but if the integration is now done over the 1.00 slice, then that is an easy calculation to make.

Note that it is important to understand that the assumption and simplification we make here is that we have a constant result over that width. Therefore, it is not as simple to multiply a width with a value and get a line element out of a surface element. That requires that the considered integration width is much (much) smaller than the full size of the tunnel, for instance, as otherwise, the accuracy of the calculation is far from sufficient.

To sum up, I hope that it is now clear that the value of 0.06m is ONLY applicable when you consider an integration over a width of 0.06m and it is NOT a generic number you need to use to consider a line element out of a surface element.

Once again, I would strongly advise discussing the way you extract and interpret your structural forces with another geotechnical or structural engineer in case of integrating structural forces of surface elements.

RE: Plaxis 3D plate output unit kN*m/m

Mikael Manuit — Mon, 17 Jan 2022 16:21:22 GMT

Thank you so much for your reply Sir Stefanos.

On the other hand, I was thinking, what if with the same cross section properties, I am going to make the length of the tunnel model=1.0m? And then for every Bending Moment I get from that model, I multiply them by 0.06m?

Do you think that would be correct to get the Bending moments of my former model which has a length of 0.06m?

Get Outlook for Android

RE: Plaxis 3D plate output unit kN*m/m

Stefanos Papavasileiou — Mon, 17 Jan 2022 14:20:18 GMT

Dear Mikael,

Thank you for the screenshot as it helps understand what you have to deal with.

So, in this case, you have a surface plate element with a width of 0.06m and you want to know what is the bending moment at specific locations. The bending moment presented in PLAXIS Output has units of kNm/m because these are the proper units for a surface element, based on the assumptions mentioned before in this post.

If you want to assume that this tunnel slice, i.e. 0.06m, is a line element you would integrate the bending moments you see, for instance, when hovering with your mouse using the Hint box (or in a cross-section), over that perpendicular direction.

Assuming that the bending moment is constant over the perpendicular direction, you would multiply the value you see with 0.06m in your case, giving units kNm. That would give you an equivalent (assuming constant) bending moment over a 0.06m slice.

RE: Plaxis 3D plate output unit kN*m/m

Mikael Manuit — Mon, 17 Jan 2022 02:05:16 GMT

I am using Plaxis 3D, and I actually have a similar problem but the difference is my model is a circular tunnel plate (dia=0.10m) and the length is less than 1 meter (0.06m.)..

The tunnel model is subjected only to the soil weight above it and I need to find the Bending Moment at specific nodes along the center of the tunnel lining (top, bottom, right and left side, nodes at 45, 135, 225, 315 degrees).

My problem is the Bending Moments, M11 and M22 are in kn-m/m.

What calculations do I need to do, to make the Bending Moments kn-m?

RE: Plaxis 3D plate output unit kN*m/m

Mikael Manuit — Sun, 16 Jan 2022 11:07:25 GMT

Sir Stefanos, what do you mean when you mentioned "integrate the bending moment over its length"? Can you please show the mathematical interpretation for this? Thanks!

RE: Plaxis 3D plate output unit kN*m/m

Dennis Waterman — Mon, 27 Sep 2021 13:54:30 GMT

Hi Sonia,

This is per foot perpendicular to the cross section. it's a distributed bending moment per meter width....

With kind regards,

Dennis Waterman

RE: Plaxis 3D plate output unit kN*m/m

Sonia Swift — Fri, 24 Sep 2021 13:43:59 GMT

Thank you - that is a huge help! And the units are still lbf ft/ft. Is that per foot in the direction perpendicular to the cross section or do I need to calculate the area under the curve drawn (like i've shown with the blue trapeziod in very poor freeform drawing).

RE: Plaxis 3D plate output unit kN*m/m

Dennis Waterman — Fri, 24 Sep 2021 07:34:25 GMT

Dear Sonia,

The reason you get multiple values for each point is because those values come from the different elements connected to that point. The forces are extrapolated from the elements for each element, so in a point that separates 2 elements one gets an extrapolated value from both elements. So in principal one should take a weighted average of those values, weighted because smaller elements are more accurate and extrapolate over a shorter distance, so the value coming from a small element has more importance than the value coming from a large element. So for example, if in a point you have 2 values v1 and v2 coming from 2 elements with surface areas A1 and A2 then the weighted value should be calculated as v = v1 * A1/(A1+A2) + v2 * A2/(A1+A2).

But you don't have to do that yourself. If you draw a cross section in 3D then in the cross section plot you can choose from the View menu the option "Use result smoothing" and then the cross section will be redrawn using the weighted values. Also in the table you will then see you still have multiple values per point, but all those values are the same now.

With kind regards,

Dennis Waterman

RE: Plaxis 3D plate output unit kN*m/m

Sonia Swift — Thu, 23 Sep 2021 21:11:22 GMT

Is there an easy way to integrate the area under the curve? I copied the table of values into Excel but I get multiple values at each point and am not sure if I should be deleting some of the duplicate values.

RE: Plaxis 3D plate output unit kN*m/m

Stefanos Papavasileiou — Fri, 17 Jul 2020 10:51:24 GMT

Dear John,

By creating a cross-section of a plate element, you can see the result, e.g. bending moment, along the cross-section length.

If you need to find the peak value and location you can use the generated plot, e.g. spotting a local peak value. For getting the bending moment at a specific location of that section indeed you integrate the results over this length.

Indeed, in general, the geotechnical engineer needs to sit with the structural engineer and identify the results one needs from another.

RE: Plaxis 3D plate output unit kN*m/m

John Chen — Tue, 14 Jul 2020 12:42:09 GMT

Hi Stefanos, Thanks for your explanation, this refreshed my mind of structural theory. I understand the output value from Plaxis 3D is the peak moment per unit width of plate strip, in local axis 1 or 2 direction. As geotechncial engineer, we do the model and output the value to structural engineer, what structural engineer mostly interested in is the peak value, which is the output result shown. When you say using the integraion of the bending moment over its length (1 or 2 directon), do you mean cut a section along plate, intergration M*dx / X to get averaged moment ? (X is the toal length)

My meomory to get moment is the shear force integration over length. Overall, I believe structural engineer is intesreted the peak moment located on a point of plate (directly read from output), rather than total net moment acting across the whole plate length. Please correct me if you see anything not right

RE: Plaxis 3D plate output unit kN*m/m

Stefanos Papavasileiou — Mon, 13 Jul 2020 13:19:25 GMT

Dear John,

The bending moments show the bending of a plate at a location in a certain direction. To get the total bending of a plate in a certain direction, you should make a cross-section at the desired location and integrate the bending moment over its length (not thickness).

Keeping in mind the simplified Grashoff method would help to understand the units adopted for the forces:

this method considers a slab/plate divided into strips of unit width into the two different directions and the forces are thus in the unit of width. Please refer also to the figures below.