Hi,
I am an aerospace engineering student, we have to build 1 meter diameter wind turbine for course project. We have an old cnc hot wire cutter, step four, of early 90's production. It works fine and properly cuts tapered wings. My problem is that our turbine blade should have a 60 degree twist between root and tip. I did couple of tests but results aren't impressive. At root I placed horizontal 25 cm airfoil, at the tip 10 cm airfoil with 60 degrees of twist around the leading edge. Leading edges of the two airfoils are aligned, trailing edge of the tip is lower.
Resulted shape does have a twist, the problem is in the half span blade vanishes. Apparently it melted because the tip airfoil was too thin. If I increase tip airfoils thickness could that do the job ?
Also, I had problem with DXF import. From an airfoil archive I downloaded dwg file, converted it to autocad 2004 dxf format, but during import only the upper surface of airfoil showed up. When opened in autocad that dxf airfoil is a single curve, it is not even splitted to top and bottom surfaces. Today I converted airfoil into autocad 2000, R12 formats, maybe that machine doesn't support newer autocad versions.

I don't have access to the cutter on weekends, I'll resume works on Monday and desperately hope to come up with the solution to this problem because presentation is on Wednesday. Sorry for long post. Any input is highly appreciated.

Ilkin

You might need to cut it in many pieces, and glue them together. Would be easier to give advice with pictures, or the dwg/dxf files.

Best way to do this is to break the job into smaller 'span' sections, and sand the final down to smooth shape. I'd assemble on a CF rod or similar.

Is this an experiment to demontstrate how useless wind turbines are? Good idea. Someone needs to.

Yes, that was my initial plan, glue together 50 sections. I bought 10mm thick plates, I was gonna cut them and stack 50 of them together, cut tapered wing, then twist while gluing. People were saying that giving twist with hot wire is not possible. I think that as long as wire begins cutting from LE and finishes at TE at both ends at the same time there should be proper twisted wing.
Span of the blade should have been 50cm, but it melted half way.
In the airfoil folder dwg file is original file that I found from some archive, then I converted it into 2004 dxf format.

Large twist won't give accurate shapes in the middle.
Picture what you would get with a simple flat section twisted 60 degrees.
Mid-span the chord is significantly reduced.
(This is exaggerated in my example since span is only twice the chord, but it demonstrates what is going on)

A good way with AutoCad to do section imports is to turn the .dat files into scripts.

Remove the name in the first line, replace it with the word "line"
Remove all blanks and add commas between the ordinates.
Save the file with a .scr extension

In Autocad type "script" at the command prompt and choose your file.
It will draw it at the current UCS 1 unit in length.
(You will have to hit enter or escape to to exit the line command at the end)

I have attached an example taken from the UIUC database
Just change the extensions by removing the .txt

Pat MacKenzie

vintage1, you don't like wind turbines at all or you just don't think that building it out of foam is a bad idea ?
Pat MacKenzie, thank you for explanation, I was able to import dat files into autocad. But with your file as well as mine airfoil there was a problem with trailing edge, last points of the bottom surface weren't connected to the TE, they were missing. Also, it doesn't give single curve, instead many little line segments that aren't connected.
Also thanks for the insight on midspan chord reduction. I though about it and it turns out even that reduction has nothing to do with span to chord ratio. That number 0.86603 is just cos(60/2). It only depends on twist angle.
I still haven't clearly understood why foam melted away in midspan. I don't think it is because chord and thickness reduces in the middle. Because, in StepFour I was giving length, width, distance from the rails for foam block, when I specified profiles for the root and tip it automatically was cutting bigger profiles to compensate for the profile reduction.
I concluded such a lesson: to cut twisted shapes, length should be less than half the distance between the rails, and the foam block should be placed as close as possible to one of the rails. Shape will get smaller in the middle, and get bigger near the rails, to avoid this foam block should be placed completely to right, or to the left to the midway point of the rails.
Ironically, this was what I did, still tip part was lost.

The many little segments can be easily turned into a polyline.
(Easy if you know how to use AutoCad)

Or you could just replace the "line" command with "pline" in the script, and you will get a poly-line

And for a clean exit from the script add the line:

close

after the last data point. Be sure to put in a single carriage return after the close

Pat MacKenzie

Yeah, that worked thank you so much.

1/. I think wind turbines are probably the greatest fraud since Ponzi.
2/. the wire speed needs to be carefully controlled when cutting thin sections, and temperature.
3/. Its easier for a one off to split into smaller sections. And sand out the humps. For wood props I split spanwise, but foam is probably easier chordwise.

Originally Posted by vintage1

1/. I think wind turbines are probably the greatest fraud since Ponzi.

I am with you on that one. I used to think that a single one of them looked kinda nice, with those long glider-like wings. But now, when they are exploiting our beautiful coast lines with hundreds, I am very much against them. The industry that make them are not environmentalists, they just jumps whatever gives most governmental funding at the moment.

Lars

Originally Posted by alxasli

----I concluded such a lesson: to cut twisted shapes, length should be less than half the distance between the rails, and the foam block should be placed as close as possible to one of the rails. Shape will get smaller in the middle, and get bigger near the rails, to avoid this foam block should be placed completely to right, or to the left to the midway point of the rails.
Ironically, this was what I did, still tip part was lost.

I think you missed whats going on. The machine is using a straight wire, and you need a curved cut. If you move the part along the table your just changing the cut, probably not getting the desired end result. That is why we have recomended cutting in smaller pieces The curves error is less, and can be sanded smooth after glueing. You can save the cut offs to align it for gluing. But to do that you need to square off the foam block, and cut each piece at its proper height, and angle.

Hi, I cut the shape I need Last time material was softer than usual, tip airfoil was so thin that after wire took its share there was nothing left at the tip. This time I used ofsetted airfoil, so that after wire cuts it's share it leaves exactly what I need.

If you take thickness and chord measurements from the middle of that blade, and compare against your design, you will likely find it differs substantially.

To get the twist to work to give a fairly accurate foil progression, you probably have to cut as sections. That is due to the distortion introduced by the large twist in your design.

A trivial example - two foils as flat plates, oriented 90 degrees apart. The intermediate foil ends up being a flat plate at 45 degrees, but sqrt(2) of the length of the ends.

Gerald

Yes Gerald, you are right about two plat foils example, but if there is a taper there is no reduction of chord in the middle other than expected. As long as tip chord is smaller than mid chord there is a gradual, linear reduction of chord from root to tip. trick is taper

You may want to play with the math on that one a bit more... It is non-linear for non-zero twist.

Gerald