Will be starting up an informal meetup starting this Saturday at 2pm (ideally meeting for consecutive weeks at the same time).
I started tearing down a lawn mower engine (Briggs & stratton 158cc) already, and just ordered the gasket kit to rebuild it. So first thing will be getting it running again, then I’m thinking of trying to [build a wooden blower] (Building a dust collector blower) (out of MDF, because free), if that works it would be a good start to a hovercraft.
Anyone interested in helping or spectating is welcome to attend, should be interesting to see how close an MDF impeller can get to 2,000+ rpm before becoming shrapnel.
Can anyone think of a good way to approach designing a backward curved centrifugal fan? Specifically how much curve, what should angle on inside vs angle on outside be, inner radius vs outer radius, etc. I could just randomly draw something up based on some pictures and the wikipedia article, but I’m thinking it would require just making a few at once and testing them all to gather any data on what works best.
Started playing around with Onshape to design the blower (starting from randomly guessed parameters), I was very impressed with how easy it was to start using, plus I can view/edit from my phone!
I downsized to an 18" diameter with sheet metal vanes, trying to reduce the mass seemed like a good idea, especially since the impeller will only be mounted directly to the crankshaft of the motor, and there are no bearings in the motor.
I remember that your outer profile was broken up into straight line segments out of OnShape and you mentioning what seemed like a pretty convoluted way to get that shape extruded etc. I just ran a quick test of essentially implementing a polar curve out of Solidworks to see if it handled it well. In the attached screenshot, I essentially converted:
r = 1 + t
Which is probably close to the form of the equation you’re using into the Cartesian parametric equations of:
x = (1+t) * cos(t)
y = (1+t) * sin(t)
and created a sketch with that. I then connected the starting point with the end point with a coincident line and had a fully constrained sketch. I assume this would be smooth and not broken up into polygon soup etc., so it’s possibly a cleaner end result:
Yeah this is how I made the spiral in onshape, a helix on a conical surface then projected into another plane. And because it was a spiral 2 feet across the line segments really showed up.
It probably would have been better in onshape to use a spline and kind of manually align the key points
TIL the name for a few spirals during a wikipedia walkabout starting at your link.
Does that book you bought say what the equation is supposed to be? It looks like the cylinder/helix is an Archimedean spiral, which coincidentally is the type I created as well.
As an aside: I’m pretty sure there’s a section in the book “Hunt for Red October” that mentions this equation can be used as a search pattern for a submarine knowing its last known location and top speed. Quite possibly only interesting to me.
Will be milling a flange adapter tomorrow, if I can find some extra 7/8 inch diameter shaft I could try to balance it by rolling that along two parallels. But mainly I’m hoping the mdf + CNC router produced perfectly balanced discs.
Just be cautious that even if it is strait & true, it can still be out of balance. I have seen wheels that were not true be balanced enough to not notice.
FWIW: Balance is probably more critical than straight-n-true from a safety perspective, it can throw itself apart almost immediately. Might be best after you’ve balanced it, to maybe test rotating it using a cordless drill so it’s turning at low speeds. If something is seriously out of whack it’ll show up then.
An 8x slow mo of the first startup, from the laser tach we tested and audio analysis it was able to hit 1,500 rpm at full throttle.
However we failed to anticipate one risk, and that was extra length on the pull coord, during a subsequent test the pull coord got sucked into the impeller causing the impeller to stop spinning sooner than the crankshaft, causing the bolt holding the impeller on to tighten itself and destroying the impeller. Turns out MDF is not very strong when subjected to extreme compression.
So will have to start over with building a new impeller, probably out of high quality 1/2" plywood instead of MDF. I am also tempted to go buy a $100 Harbor Freight predator engine, which is 6.5hp, 2-3x more powerful than the current lawn mower engine.
