Loads generated by RWIND and interface with RFEM

Hello everyone.

I have been doing some simulations on RWIND 3 and I want to check some of the results by hand. To perform the verifications correctly, I would like to know the answers to these two questions regarding how RWind works :

1 - how does RWIND convert the velocities generated by the wind tunnel into pressures (what method/ formulas uses, wich coefficients it chooses, etc...)?

2 - does RFEM 6 apply some coefficient to the wind loads generated by RWind? Because I tried to do some verifications by hand and the values don´t seem to be the same.

As a note, I am using the Eurocodes as standard for the calculations.

Best regards,

Pedro

Hello Pedro,
Thanks very much for your interesting questions.

1) How RWIND converts velocities into pressures

RWIND uses a numerical CFD (Computational Fluid Dynamics) approach to simulate wind flow around structures in a virtual wind tunnel. The airflow is governed by the Navier–Stokes equations together with an appropriate turbulence model (typically RANS). In this framework, pressure and velocity are not derived from each other using a single formula; instead, they are solved simultaneously as part of the governing equations of the flow.

The commonly known expression:

represents the dynamic pressure, which is primarily a reference quantity used for normalization (for example in pressure coefficients). It is not the actual pressure acting on the building surface.

RWIND directly computes:

  • Local static surface pressures on the geometry

  • Including effects of stagnation, acceleration, separation, and turbulence

  • Based on the full CFD solution

*How pressure coefficients (Cp) are computed in RWIND

When RWIND reports Cp, it uses the standard definition:

where:

  • p = local surface pressure obtained from the CFD solution

  • pref​ = reference (far-field) pressure

  • Uref​ = reference inlet velocity of the CFD model (typically the free-stream velocity of the primary model, but editable by the user)

  • ρ = air density

It is important to note that:

  • The reference velocity and height used in RWIND may differ from those defined in Eurocodes,

  • Therefore, CFD-based Cp values cannot be directly compared to tabulated Eurocode Cp values without consistent normalization,

  • Eurocode Cp values are area-averaged and conservative, while RWIND provides high-resolution local pressures on the exact geometry.

2) Does RFEM 6 apply coefficients to RWIND loads?

RFEM 6 does not modify or scale the wind pressures imported from RWIND. The pressures are transferred as externally defined surface loads.

However, during the structural design stage, RFEM naturally applies:

  • Partial safety factors

  • Load combinations

  • Design situations
    in accordance with the selected standard (e.g., Eurocodes).

As a result, discrepancies between hand calculations and numerical results typically arise from:

  • Different reference wind velocities or heights

  • Comparison of CFD mean pressures with Eurocode conservative peak values

  • Differences in terrain roughness or inlet profile definitions

  • Comparing characteristic RWIND loads with design values in RFEM

Exact point-by-point agreement between Eurocode tables and CFD results is neither expected nor methodologically correct. What should be consistent is the global force level and overall trends.


Practical recommendation for verification

When verifying RWIND results manually:

  • Compare characteristic values with characteristic values

  • Use the same reference dynamic pressure

  • Do not directly compare CFD Cp with Eurocode Cp without proper conversion

  • Carefully check reference height, wind profile, and terrain parameters

*In addition, there are insightful validation examples related to the Eurocodes and other standards on our website, which are highly recommended for review at the following link:

:backhand_index_pointing_right: https://www.dlubal.com/en/downloads-and-information/examples-and-tutorials/verification-examples?q="Mahyar%20Kazemian,%20M.Sc."&f=_

2 Likes

Hello @mahyar.kazemian

Thank you for the reply and the information you gave :grinning_face:

After reading your answer I have some extra questions:

1 - You mencioned that the pressures determined by RWIND are imported as surface loads. Is there a way to see these loads and check their values?

2 - I will explain the methodology I used to check the values: I checked the value of the moments generated by the load case (not load combination) containing the RWIND results (see the values bellow).

With the values in hand, I used teh general equation for moments (pl^2/12) to verify the load on the purlin with the highest load. And after the calculation was complete, it showed me a value of the wind load higher than the value of the pressure generated by RWIND (see images bellow)

What I want to ask you is if my methodology is right, and, if it isn´t, could you tell me how to check the values propperly?

Best regards,

Pedro

You're welcome Pedro :slightly_smiling_face:

1) Can RWIND surface loads be viewed and checked in RFEM?

Yes, you can find those pressure values in both RWIND (by using point probes) and RFEM (by activating the value on the surface) as follows:

Useful link for more info:
:backhand_index_pointing_right: https://www.dlubal.com/de/downloads-und-infos/dokumente/online-handbuecher/rfem-6/001107


2) Important point about using the formula M=pl2/12

That formula can be valid under specific assumptions, such as:

  • Uniform pressure

  • One-dimensional load transfer

  • Simple beam behavior

  • No load redistribution

  • No plate or shell stiffness

  • No spatial pressure gradients

In your case:

  • The pressure field is highly non-uniform (CFD result with local peaks and gradients)

  • The load is transferred through roof panels + purlins + frame interaction

  • The FE model redistributes forces through plate/membrane action

  • Supports and stiffness variations affect internal force paths

  • RWIND applies pressures to surface elements, not as equivalent uniform beam loads


3) How to verify RWIND loads correctly (recommended methods)

If you want a physically meaningful verification, use one of these approaches instead:

:white_check_mark: Method A – Direct pressure integration

  1. Export nodal pressures (or surface pressures) from RWIND or RFEM.

  2. Integrate pressure over the tributary area of a purlin:

  3. Convert that force into an equivalent line load and compare.

This keeps consistency with the actual CFD field.


:white_check_mark: Method B – Resultant force equilibrium check

Check global equilibrium:

  • Sum of all applied RWIND forces on the roof
    vs.

  • Total reactions at supports

These should match (within numerical tolerance).


:white_check_mark: Method C – Compare pressure levels, not internal forces

Instead of checking moments, compare:

  • RWIND pressure magnitudes (kN/m²)

  • Against:

    • Expected Cp ranges

    • Eurocode external pressure envelopes

    • Wind tunnel literature for similar geometries

This validates the aerodynamic realism, which is the correct level of validation.


4) Why your calculated pressure can be higher than RWIND

This happens because:

  • The structure redistributes load internally

  • Local stiffness attracts more force

  • Your back-calculation assumes uniform loading and pure beam behavior

  • Peak moments often correspond to local pressure concentrations, not average pressure

Please let me know if you have more questions :slightly_smiling_face:

3 Likes