What are the best practical ways to model multistorey CLT or timber frame building, specifically in regards to wall anchors? I have to (in most cases) do a modal analysis. I can model supports, line hinges and springs as non-linear to take stiffnesses and behaviour in account, but modal analysis is linear analysis and non-linearity will not be taken into account.
How do you do it in practice? Is there a way Dlubal recommends?
If I decide to ignore non-linearity and model everything just with line hinges and consider only shear stiffnesses, how to get uplift forces faster? Right now I can read vertical forces (uplift and compression) from result diagrams and "integrate" them on each end but that is very time consuming for each wall.
Hi wolfram,
Thank you for your question. Here is how we would approach both points.
1. Wall anchors in a dynamic (linear) analysis
You're right: the modal analysis and the response spectrum analysis are linear, so nonlinear supports, line hinges or springs (for example, failure under tension) are not taken into account. In practice, we suggest this approach:
- Response spectrum analysis (linear): Model the line hinge between the wall and the floor below with a realistic shear stiffness in the wall plane (u_x). Make it as stiff as possible in the vertical direction (u_y). This gets you closest to the realistic deformation, and you can still use the response spectrum method.
- Pushover analysis (nonlinear): If you want to capture the real anchor behaviour, including the hinge failing under tension (uplift), you need a nonlinear analysis, such as a pushover analysis.
To show the effect, we compared several variants of a simple two-storey wall under the same horizontal load:
linehinges.rf6 (1.1 MB)
| Variant |
Line hinge |
Max. deformation |
| A – linear |
u_x = 2000 kN/m, u_y = 2000 kN/m |
40.1 mm |
| B – linear |
u_x = 2000 kN/m, u_y = 20000 kN/m |
11.7 mm |
| C – linear |
u_x = 2000 kN/m, u_y = rigid |
8.5 mm |
| D – nonlinear (reference) |
u_x = 2000 kN/m, u_y = rigid, failure under tension |
8.8 mm |
A linear spring has the same stiffness in tension and compression, so a soft u_y (A) greatly overestimates the deformation. Increasing u_y is a simplification, but it brings the result close to the nonlinear reference (B/C vs. D). That makes it a good basis for the response spectrum analysis.
2. Getting uplift forces faster – smooth ranges in result diagrams
You don't need to integrate the vertical forces by hand. The Result Diagrams window has a Smooth Ranges function that does this for you:
- Open the result diagram of the line hinge (Results on Lines → Line Hinges → Forces → v_y).
- Click the Smooth Ranges button in the toolbar.
- Define a range at each end of the wall (for example, Start 0.000 / End 0.480 m and Start 1.607 / End 2.000 m), and select "Constant" as the distribution.
- Under "Use for Results", select Line Hinges → Forces → v_y.
For each range, RFEM then shows the resultant force Σ (e.g. −33.48 kN and 25.81 kN), the mean value, its location x and the range length. These are the tension (uplift) and compression forces at the wall ends that you need for the anchor design.
You can find more details in our FAQ:
If you have any further questions, please let me know.
Best regards
Florian Heinrich
