Modeling Challenge: Propeller
I’ve got to admit, I had a bit of difficulty getting a propeller that was even recognizable. This challenge may be more difficult than it looks. Consumer product, “artsy surfacing”, is easy enough, because no one knows if you are faking it. There is no such thing as “precision” in that kind of modeling. As long as it looks good, you’re ok. I’m not trying to put down that kind of work, because I do plenty of it.
The difference with technical surfacing is that you’ve got to satisfy the asthetic sense, but only after you’ve satisfied the math. I didn’t really use any math for this model, but people who do make propellers and other air or fluid foil shapes do need to be technically precise. A recent boat hull project brought some of this to my attention.
So, the challenge is to model a propeller. You can model a plane or boat propeller, or some other type, such as a pump or compressor impeller – any sort of turbomachinery type device will do. I’ll go through each model and make comments based on a few things:
- starting reference data if any
- quality of sketches
- inventiveness in SolidWorks surfacing techniques
- does it actually look like a propeller?
- surprise me
Anyway, my comments may not come immediately, this is a busy week for me. Send in your models to my email address, matt at dezignstuff dot com. I’m really looking forward to this. Technical surfacing is very difficult.
=============================
Entry #1 is from Matt Sass. Matt swept an airfoil shape along a straight line, using the “Twist Along Path” option, twisting 60 degrees. Thanks for that entry! The first step is the most courageous. Click the image to download the part. Keep ‘em coming!
=============================
Entry #2 is from Garrett Brooks. This is a very nice job. Garret searched the web and found a nice sketch picture, then traced the various profiles and used Derived Sketches to place the sections where they should go, then made a Boundary surface with several guide curves. The ends of the blades are capped off with Fill surfaces. This one is nicely done. You owe it to yourself to download it and check it out.
=============================
Entry #3 is from Rick McWilliams. This is another very nicely done model. Rick made a boat propeller instead of a plane prop. Rick used a different technique on this one. Because the blades take up a bigger wrap angle around the prop, it made most sense to lay the lofts out using cylindrical cross sections instead of flat cross sections. The required flat sketches that were wrapped onto cylindrical faces. Then rick lofted through the wrapped edges, and lofted to a point to cap off the blade. Very nicely done. This one is saved in an older version ofSolidWorks. Don’t know which one, but it is at least earlier than 2009.
=============================
Entry #4 is also from Rick McWilliams. This is an assembly of a cooling fan. Download Ricks assembly and parts and see how he made the blade.
=============================
Entry #5 is from Dave Yeagly. Dave took the simple but effective route with this one. He did a two line loft to get a twisted shape, then trimmed it to get the prop blade shape.
=============================
Entry #6 is from Robert Veronneau of Hammerhead Aviation. Robert had a P-51 Mustang propeller blade sitting around the office, and decided to digitize it. It looks like he used DezignWorks with a romer arm. Robert says “The sketches (Splines) have not beed cleaned up yet, so the entire model is a bit rough.” Yeah, looks REALLY rough, Robert. Actually it looks great. Thanks for sending this model!
=============================
Entry #7 came from Garrett Brooks. This is a nicely done surface model. Download the model and check it out.
=============================
Entry #8 is from Mark Reader. Finished in Solidworks, but done in MoI. Here is what Mark had to say:
Wow, great challenge…… First off, I intended to do this model completely in SW but it became too frustrating and so it was back to MoI. Not trying to be a fanboy or sell it, but it is a fun program to work with. Like Garrett I spent time searching for some drawings but I like the one he found better. My approach was similar to his but I found it difficult to get rid of the waves, dips & bulges I was getting. You can see some in his model too if you look down from the tip of the prop. They’re there, but not that bad compared to what I was getting.
Attached are 2 props done in MoI, brought into SW and rendered. 1 is semi-realistic and the other just for fun. The import into SW wasn’t as clean as it’s been for me before. I got several “extra” surfaces so there was some clean-up needed. I also didn’t pay much attention to size/scale. If time allowed I wanted to work on the wing tips a bit more but…..
Entry #11 is from Chris Huntley. Chris made a turbine blade. Here is what Chris had to say about this model:Another hack’n’slash attempt, somewhat late and maybe not what you were looking for…The attached is an eyeballed turbine blade, the airfoil modelled using a boundary surface. The section sketches are derived from a master. What I didn’t know was that you could drag and scale a proportional spline in a derived sketch…. Haven’t explored this any further.
No idea how near realistic it is (not very at a guess).
The last time I attempted to model one of these was using SWX 98+ when I had to try to reconstruct a blade from a drawing (coord sections) to use as a scaled calibration target. This was for a set of tip clearance probes a former employer used to make. It was completely impossible…
Keep up the good work,
it’s a nice blog.
could you tell me how to draw 3d propeller blade in SolidWork step by step? or do you have a tutorial?? thanks
Nice propeller Paul. Solidworks seems to do a good loft when the sections are wrapped on cylinders. The leading edge and trailing edge guide curves are splines constrained by the specified sections and a tip slope. Solidworks does an amazing fillet at the hub. Lofts are not necessarily trustworthy in that they sometimes ignore profiles. Boundary surfaces provide more certain shape control.
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2010/08/GL_BladeDeveloped_Closup_Aug2010.jpg[/img]
Oops a year late as usual.
OK, all good fun making models look like propellers BUT, as Walter says they have had to write their own program to develop the correct foil curves. There are programs like PropCad for purchase. Most companies however develop their own software.
I was doing waterjet prop design with a software curve set and SWX 2004. The issues where firstly. how to get a surface to wrap around a NACA foil without rippling behind the leading and trailing edges. We never solved that in SWX.
The second was getting 10 (we had to split the curves in 2 and apply a face to each side to overcome the first problem) curve sets into the model space without breaking a finger clicking. It took about an hour of tedium from Excel out wards.
The solution was provided by VXCAD which had a macro that imported the full foil cross sections as one CSV file AND automatically applied the surfaces AND closed the ends to produce a solid blade in 4 mouse clicks or about minute. These were ripple free same data. I ripple free. I subsequently became a VX user and still am.
Walter, the way we solve the hub to root problem is to include the radial location geometry as part of the foil curves. e.g the cross sections are wrapped around a circ as set for hub radius. We think this is the right way to think about rotors anyway.
FYI the blade shown is a dummy data but uses a NACA (63?)foil is warped and mod’ed in various ways to try and achieve consistent mass flow across the foil.
The real rotors (more blades than this) actually worked really well and exceeded cavitation expectations by about 30% which we thought was OK for a first effort.
A couple of screen shots – showing curves used and wrap built in plus finished example. I just rehashed an existing item rather than build an aero prop. Hope you don’t mind.
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2010/08/GL_BladeDevelopWireframe_Aug2010.jpg[/img]
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2010/08/GL_BladeDevelopEdgeHighlight_Aug2010.jpg[/img]
[img]http://www.dezignstuff.com/blog/wp-content/uploads/2010/08/GL_BladeDeveloped_2Bl_Aug2010.jpg[/img]
We are producing aircraft propellers since 1985 and use some custom made software. But the hub blade connection is still problematic. It is either a lot of work to blend these two areas perfectly or the programming result leaves the root area unfeasable, so it needs manual work.
I can provide a perfect curbed blade in 3D data file with any pitch and stile in 5 minutes, but it takes me several hours to fix it perfectly to a hub.
I would be glad to hear about any easy methods that I coudl apply without learning a new software, apart from Intellicad.
Dear Matt,
Can you provide a tutorial on how to model the propeller for entry #13? I am currently doing an engineering project for my masters degree incorporating the structural analysis of an ROV thruster.
Thank you
Hai friend it real good modeling work.It was really helpfull to me.I would like to have a wide look at your models can u send me those models to my E-mail id:sp_prathap@hotmail.com.
with regards
prathap.S
Just wanted to check to see if you got my model that I sent a few days ago for the prop challenge. I like the website, and I like the little challenges, I have picked up a few ideas for the future.
Adam
Hello Matt, very interesting blog!
Hello everybody!
I was having some time today, and I’ve started to browse the SolidWorks forums. I’ve found your posts and then this very interesting design challenge.
I have done a propeller in solids (the one strictly in surfaces is not “working” yet). How can I send you my SW model?
Thanks,
Adrian
Alex_SW
Send it to my email – matt at dezignstuff dot com
Some thoughts on engineering from the largest model airplane prop maker
http://www.apcprop.com/v/Engineering/engineering_design.html#cad
My thought was to use several helix’s of different radius, but of the same pitch, to yield points where I could attach airfoil sketches to getter a proper “true pitch” prop, but I haven’t made time for it yet. You can import NACA airfoils (100 points in x,y format) by using the “curves through x,y,z” function. Just put a column of 0’s for z. Chord will be 1 unit, then scale.
I would never try to model this, I’d just hire you.
Don’t modern airplane props and boat props use airfoil data? I like how modern airplane props have inflight adjustable pitch.
Thanks,
Devon