# pitch and torque spikes on variations

**URL:** <https://forums.nlr.gov/t/pitch-and-torque-spikes-on-variations/1762>\
**Category:** Computer-Aided Engineering Software Tools\
**Created:** [January 9, 2018, 7:08pm UTC](https://forums.nlr.gov/t/pitch-and-torque-spikes-on-variations/1762 "2018-01-09T19:08:34Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Leonardo.Lopez](https://avatars.discourse-cdn.com/v4/letter/l/ecb155/32.png) [@Leonardo.Lopez](https://forums.nlr.gov/u/Leonardo.Lopez)\
**Post date:** [January 9, 2018, 7:08pm UTC](https://forums.nlr.gov/t/pitch-and-torque-spikes-on-variations/1762/1 "2018-01-09T19:08:34Z")

</div>

Hi!  
So, I’ve been “living” with this issue for a while now, at first I simply accepted it and manage to reduce its effect but now I’d like to be sure if is creating some bad dynamics in my system.

What happens is that when I send a different reference to pitch and torque to FAST in SIMULINK it gets some spikes, imagem below will clarify this.  
It is possible to reduce the spike by adding a filter prior to sending the value to FAST, but it will still be there, but mitigated.

![Sem título.png](https://europe1.discourse-cdn.com/nrel/original/2X/b/bdefa04f4bc9d21ba4b44d0e0db1a99cce0cd4fd.png)

I saw a close problem to this in a different topic and, as the below quote, tried the following solutions without success:

> [@Jason.Jonkman](#):
>
> Dear René,
> 
> I would not expect you to receive spikes in the edgewise bending moment.
> 
> I would suggest you try a few things i.e. do the spikes go away if you:
> 
> - Disable unsteady airfoil aerodynamics (AFAeroMod = 1)?
> - Disable your self-written Bladed-style DLL controller?
> - Reduce the FAST time step (DT)?
> - Add a FAST correction step (NumCrctn \> 0)?
> 
> Best regards,

---

<div class="post-metadata">

**Author:** ![Jason.Jonkman](https://dub1.discourse-cdn.com/nrel/user_avatar/forums.nlr.gov/jason.jonkman/32/3086_2.png) [@Jason.Jonkman](https://forums.nlr.gov/u/Jason.Jonkman)\
**Post date:** [January 10, 2018, 2:05pm UTC](https://forums.nlr.gov/t/pitch-and-torque-spikes-on-variations/1762/2 "2018-01-10T14:05:23Z")

</div>

Dear Leanardo,

The FAST S-Function will use extrapolated values of the inputs from Simulink when advancing from time step n to n+1 (linear or quadratic extrapolation, depending on your setting of InterpOrder). This extrapolation may cause problems (e.g. the spikes you are seeing) if the pitch-angle command is not a smooth function e.g .when stepping up the pitch angle as you are. I would suggest ramping up the pitch angle with a smooth function to avoid problems with the extrapolation (preferably with a function whose first derivative is defined e.g. an S-shaped curve would be better than a step or a linear ramp between distinct pitch angles).

Best regards,

---

<div class="post-metadata">

**Author:** ![Leonardo.Lopez](https://avatars.discourse-cdn.com/v4/letter/l/ecb155/32.png) [@Leonardo.Lopez](https://forums.nlr.gov/u/Leonardo.Lopez)\
**Post date:** [January 10, 2018, 3:27pm UTC](https://forums.nlr.gov/t/pitch-and-torque-spikes-on-variations/1762/3 "2018-01-10T15:27:48Z")

</div>

> [@Jason.Jonkman](#):
>
> Dear Leanardo,
> 
> The FAST S-Function will use extrapolated values of the inputs from Simulink when advancing from time step n to n+1 (linear or quadratic extrapolation, depending on your setting of InterpOrder). This extrapolation may cause problems (e.g. the spikes you are seeing) if the pitch-angle command is not a smooth function e.g .when stepping up the pitch angle as you are. I would suggest ramping up the pitch angle with a smooth function to avoid problems with the extrapolation (preferably with a function whose first derivative is defined e.g. an S-shaped curve would be better than a step or a linear ramp between distinct pitch angles).
> 
> Best regards,

Thank you again for your accurate and quick response!

Indeed by adding an s-shaped smoother fixed the issue.

For anyone having the same problem:

Add a transfer function of the form “Y(s) = H(s)_X(s) where H(s) = K_wo^2/(s^2 + 2_zeta_wo\*s + wo^2). Here you define zeta = 1.0. Then the s-shape is retained for any K and wo value. Note that X(s) has already been hit by a ramp. In matlab or any other tools, linear ramp and 2nd lag are standard blocks.”

source: [stackoverflow.com/questions/188 … n-simulink](https://stackoverflow.com/questions/18871488/how-to-smooth-rectangular-signal-with-high-order-rate-limiter-in-simulink)

All you gotta do is make that TF fast enough by increasing “wo”