MatLab ski calculation

For discussions related to ski/snowboard construction/design methods and techniques.

Moderators: Head Monkey, kelvin, bigKam, skidesmond, chrismp

Buuk
Posts: 150
Joined: Wed Apr 26, 2006 2:49 am
Location: The Netherlands

MatLab ski calculation

Post by Buuk »

Hi there,

After a great post of Bambi a while ago about an Excel sheet to calculate the stiffness / deflection of a ski I was inspired to work on this as well.

Last weeks in my free time I have been working on some MatLab code to show the following:

- Ski contour
- Ski profile
- 3D view of the ski contour and profile
- Binding position estimation (and percentage of area in front and behind binding)
- Ski bending stiffness (E*I)
- Ski deflection
- Slope of deflection

Well, this might look about the same as the Excel sheet of Bambi, but now the differences:
- Stiffness is calculated using parallel axis rule (don't know if this is the correct English word...) and the properties of 3 fiberglass layers above and 3 layers below the core can be set independently.
- I also used Finite Element Analysis to calculate the deflection and slope of the ski. Using this method it's easy to take a look at different loading situations.

Below some pics of the result (sorry for the huge size!)


Image
The black rectangle in the ski boot position plot is the boot :)

Image
In the plot of the deflection the green surface is the load at the ski, in this case a varying distributed load with a max of -0.6Nmm at the mounting area.

Later I hope to include the base, edges, sidewalls and topsheet as well.

Please let me know what you think about this!!
I will post the MatLab files soon!


Buuk
Last edited by Buuk on Wed May 04, 2011 1:26 am, edited 1 time in total.
Make things as simple as possible, but not too simple
User avatar
bigKam
Site Admin
Posts: 538
Joined: Tue Feb 01, 2005 5:15 pm
Location: Park City, Utah
Contact:

Post by bigKam »

Buuk:

wow! that's a great start! i'm glad you're using Matlab. for those who are unfamiliar with the software, look here: http:///www.mathworks.com. it's a great piece of software to solve problems numerically. there's even an open source version call Octave, which i've used from time-to-time...

anyway, i'm intrigued by your result... how was the FEA integrated? i can't wait to the addition of edges, topsheet, etc...
Buuk
Posts: 150
Joined: Wed Apr 26, 2006 2:49 am
Location: The Netherlands

Post by Buuk »

We use Matlab a lot at university, it's a great tool! I'm a fond of open source software as well, but I prefer Matlab because I am used to it and I can get it for free at my university. Another alternative open source program is Scilab http://www.scilab.org/

For the FEA model I simply used a 1D beam model. I divided the ski in a number of elements similar to the length in mm. For each element a stiffness matrix is calculated using the properties (E, I) of the element. This is combined for all elements in one huge stiffness matrix. Now nodal forces (and moments) have to be inserted and a vector is created of this. I use the constraints of a simply supported beam. Now the system K*u= F can be solved for u for all free nodes. So it's very simple...

Buuk
Make things as simple as possible, but not too simple
User avatar
Orion
Posts: 10
Joined: Mon Jan 15, 2007 8:05 am
Location: S. Seattle

Post by Orion »

Buuk-

Pretty cool. I remember when I had access to Matlab for free in school ...

Did you create your own code for the FEM in MATLAB, or are you using another application like Nastran, Ansys, Cosmos ...? If you are using a canned application, how are you integrating the results into Matlab?

Also, for the deflection and slope angle plots, should your load label be N/mm. I'm guessing you meen a distrbuted load not a moment. What is the total load you applied to get the deflection in your plot? ~700N?

Orion
Buuk
Posts: 150
Joined: Wed Apr 26, 2006 2:49 am
Location: The Netherlands

Post by Buuk »

I did write my own simple FE code in Matlab, I did build it as you can read above.

About the load, the load is a line load, so it should be in N/mm. This is only an error in the text in the uploaded pic. Thanks! I did just put some load on the ski to test it. See te loads below (also in N/mm):

Ptipbegin = 0;
Ptipend = -0.3;
Ptapertipbegin = -0.3;
Ptapertipend = -0.6;
Pmountbegin = -0.6;
Pmountend = -0.6;
Ptapertailbegin = -0.6;
Ptapertailend = -0.3;
Ptailbegin = -0.3;
Ptailend = 0;

In the picture below you can see the result of a load of 400 N distributed over the mounting area.


You can find the Matlab code at this location:

[UPDATED LINK] http://www.lucverhamme.nl/skibuilding/s ... ations.zip [UPDATED LINK]

You can adjust the variables in the file 'skivariables.m'. After saving run the file 'runme.m' and the results should show up.

The files are still a little messy, I need to clean them up and write everything more compact where possible. Also shear forces, camber, legends, etc have to be added. But that's for later...
I will post any future version as soon as it's ready.

I hope you like it and please give me some feedback about possible errors.



Buuk

Image
Last edited by Buuk on Wed May 04, 2011 1:24 am, edited 1 time in total.
Make things as simple as possible, but not too simple
Buuk
Posts: 150
Joined: Wed Apr 26, 2006 2:49 am
Location: The Netherlands

Post by Buuk »

Yesterday I added base material, edges, sidewalls and deflection due to shear to my MATLAB ski model (and made the code faster). As expected this makes the model a lot stiffer. I also noticed that the shear is neglectible compared to the bending deflection.

I will post the new MATLAB files later on.

Below a picture that shows the ski bending stiffness for the model with and without base, edges and sidewalls. As you can see, the stiffness increases more at the smaller part of the ski, because there the edges and sidewalls have more infuence on the stiffness of that section.
Image

Buuk
Last edited by Buuk on Wed May 04, 2011 1:28 am, edited 1 time in total.
Make things as simple as possible, but not too simple
hiona
Posts: 3
Joined: Thu Mar 01, 2007 12:10 am
Location: Voiron, French Alps, France

Ski calculation

Post by hiona »

Hi,

My first job was to design and calculate skis for the first ski at a major ski brand company at the end of the 80's...
I built a "home made" software implemented on PC's, Mac's and CAD stations. It still works very well.
You have to enter data : side cut, a first draft of thickness distribution, materials (you only have to choose in a existing library), and ski stiffnesses that are the objectif to reach.
The software finds automatically the right thickness distribution according to the ski stiffnesses you asked.
This software was written in fortran for PC's and it works properly.
I don't sell it. I can give it.
Your software seems to work well too.
J Lx
Voiron, FR
WSM
Posts: 3
Joined: Tue Dec 21, 2010 7:30 am

Post by WSM »

Buuk,
I cant seem to access the file via the link you provided... would love to try out the program!
cheers
twizzstyle
Posts: 2204
Joined: Tue Mar 07, 2006 8:25 pm
Location: Kenmore, Wa USA

Post by twizzstyle »

Four year old thread. Good luck.
Buuk
Posts: 150
Joined: Wed Apr 26, 2006 2:49 am
Location: The Netherlands

Post by Buuk »

I will re-post the link soon. Didn't have time to search for the files on my external HD...

Cheers, Buuk
Make things as simple as possible, but not too simple
Buuk
Posts: 150
Joined: Wed Apr 26, 2006 2:49 am
Location: The Netherlands

Post by Buuk »

Finally I found the files of the ski calculations...
Check them out at: http://www.lucverhamme.nl/skibuilding/s ... ations.zip

The file contains different Matlab m-files and I once started with the creation of a graphical user interface but didn't complete this yet.

I would really appreciate it if any updates on the files will be shared on this forum to support other skibuilders in getting more insight in ski-stiffness patterns.

Cheers, Buuk

p.s.: Links and pics are updated in my previous posts...
Make things as simple as possible, but not too simple
twizzstyle
Posts: 2204
Joined: Tue Mar 07, 2006 8:25 pm
Location: Kenmore, Wa USA

Post by twizzstyle »

Nice! Gonna have to dust off my Matlab CD and play around. Thanks for this!
ggardner90
Posts: 146
Joined: Mon Jun 13, 2011 9:02 am
Location: Salt Lake
Contact:

Post by ggardner90 »

Thanks so much, i have been learning how to use matlab just so that i can use this program to better design my skis. this has been such a huge help because solid works is not as good with the simulations of wood because of the variable of moisture. and i am not that good at it.

just wondering how many of you guys are engineers?

im an accounting and econ major. so alot of this stuff is new to me
twizzstyle
Posts: 2204
Joined: Tue Mar 07, 2006 8:25 pm
Location: Kenmore, Wa USA

Post by twizzstyle »

Aeronautical engineer here, I am a flight test engineer by profession.
ggardner90
Posts: 146
Joined: Mon Jun 13, 2011 9:02 am
Location: Salt Lake
Contact:

Post by ggardner90 »

That is sweet. i want to take some engineering courses at the local cc because i am not on the same page as most of the engineers i talk to and i really want to learn the stuff they know
Post Reply