Design and build side shaft hit and miss engine from bar stock

[Brian Rupnow]

For anybody that didn't quite get it about the gas line----These small engines will lift fuel at least 1" up from a tank, due to the vacuum created by the venturi in the carburetor. However, when the engine goes into it's "miss" cycle and no air is being pulled in thru the carburetor, the vacuum disappears---and all the gas in the line will run back down into the tank due to gravity. Then when the engine has slowed down enough to "hit" again, it can't because the gas has all ran down into the tank. To prevent that from happening, a check valve is installed in the vertical portion of the gas line. This check valve allows gasoline to flow from the tank towards the carburetor, but as soon as the gas tries to flow back down into the tank, the orange colored ball "seats" in the conical portion of line just below it and won't let the gas run back to the tank. When the engine slows down enough to want to "hit" again, it has a supply of gas right there waiting for it.

[crueby]

Great info about the gas flow!

[Brian Rupnow]

The time has come, the Walrus said, to speak of many things---And the only two major parts left are the cylinder head and the face cam. I will stick with something that I know relatively well, and make the cylinder head.

[Brian Rupnow]

Brian, you missed George's comment that the "lobe" of the can should be radial to the axis not straight across. If you imagine that grinding stone on the other forum doing the cutting while the cam is rotated horizontally it won't give the cut you have shown. You also look to have a very short duration.

Jason--seeing as it was you who posted this, it appears that you must have certain knowledge about face cams. My knowledge extends to the fact that a face cam revolves at 1/2 the speed of the crankshaft, same as a conventional cam. This being the case, the "flat" portion of the face cam should correspond to the "flat" area of a lobe cam in terms of degrees. On a conventional "lobe" cam, the rise is very gradual to the peak, and again very gradual on the downhill side of the cam. The "dwell" at the apex of a lobe style cam is really quite small.  On a face cam, it appears that the "rise" is very abrupt and of short duration, the "fall" is very abrupt and again of short duration. However, the amount of "dwell" (that area where a conventional lobe is at it's "highest lift" ) is quite long on a face cam. I have googled "design of face cams" without a great deal of success.  I know that there are "lap" and "lead" considerations on a conventional cam, which have to do with the abruptness of opening and closing the valve, which again has to do with the flank radius on the lobe. I'm not sure how that cross references to a face cam. Any wisdom you could share would be greatly appreciated.---Brian

[Jasonb]

As the follower is a larger diameter the resulting movement of the rocker will not be as direct as a typical pushrod at right angles to a conventional cam so the opening and closing will be a smoother and of longer duration that it looks on the cam.

A slow reving hit and miss engine is not going to be anywhere near as critical as a high performance engine where valve events are far more critical and if you take a look at a few H&M engines that use conventional cams they are very much "flat topped" compared to a high lift performance cam particularly the ones with Ignitors.

best to plot out the exhaust valve movement with your cad and get the point that the exhaust starts to open and is fully closed the same as the engine you have taken your head design from and not worry so much about the fact the valve is not moving constantly.

[Brian Rupnow]

This is my first shot at designing a face cam. On the right you can see the conventional can for the "odds and ends" hit and miss engine. On the left is the face cam for the sideshaft engine I am designing.--Opinions please.--Brian

[Jasonb]

Brian, as I said above the larger diameter roller will not follow the face of the cam exactly so your 18.75 won't be correct. I expect it to start opening earlier and close later.


Not sure of the sizesyou are using but I have stetched it out assuming the cam is 1" OD, 1/8" thick running surface so inner edge of the "cam" is 3/4" diameter. 1/16" lift with 1/32" radius corner and fillet. 1/2" dia follower roller,

If you then plot this as a straight line along the inner edge you get a total cam length of 0.75 x pi or 2.356" and 90degree duration would be a 1/4 of that or 0.589"

You can see that the roller will start to be affected before the point where you have picked up your 18.75deg so total affected length is 0.941 or 143.8degrees

So actual cam is 26.9deg lift duration, 90degree fully open and 26.9 deg drop

larger roller would increase the lift and drop times.

[Brian Rupnow]

A good start was made on the cylinder head today. All of the lathe work is now finished, and the next move will be over to the milling machine.--A little story here--I have two metal suppliers in Barrie. One is a fab shop that's been around forever, they do excellent work, and they sell material over the counter. My other supplier is a young guy who does nothing except warehousing and selling material cut to length. I wanted a 1 foot length of 2 1/4" diameter aluminum. The young guy is shut down for Christmas holiday. I went to the fab shop today, and asked "How much for a 1 foot length"? The answer was $17.00. Okay, I can live with that. BUT---Unless I can find a "short" out by the rack, there is going to be a fifteen minute minimum labor charge if somebody has to cut it for me. The fifteen minute labor charge would be $18.00 The labor charge would be more than the material. I said "Screw it!" and came home and whittled my cylinder head out of a piece of plate.

[Brian Rupnow]

My initial intent was to use a bearing I have kicking around as a cam follower. That bearing has an outside diameter of 0.867". My largest endmill is  1" diameter. Since I will be cutting the ramp on the cam with my largest endmill, this more or less dictates that my cam follower diameter must be. I should be okay. The rad of my endmill is 1/2". The rad of my bearing is .4335".

[gbritnell]

Brian,
As Jason showed in his sketch this is what you're trying to achieve.
This will basically be the difference between a flat tappet cam and lifter and a roller cam and lifter.
I'm attaching a sketch I made. I'ts just a general layout looking at the face of a cam disc with a roller sitting on it.
The top view shows the rise on the cam with the blue roller just touching it. The red line shows the tangent point of roller to cam rise.
As a roller cam (automobile type) has a different shape than a flat tappet cam with the same timing numbers because of the point of tangency this is the same for your hit and miss engine.
The front view shows the red tangent line and how it would look on the cam. It looks skewed because the roller is on the same centerline as the cam disc so the tangent point is offset but parallel to the axis of the roller.
The next view to the right shows a rotation of 90 degrees (180 crank) and the roller would be down on the flank of the disc.
If the lobe or high spot, whatever you want to call it, is constructed as a radian (green line) then the innermost part of the circle would hit first because it's not parallel with the axis of the roller. This cam would work but over time the timing would change because with the roller wheel hitting that tiny contact area would wear it away thus reducing open duration. The tiny contact area would also wear a notch in the roller wheel.
I have to apologize for the previous sketch. After looking at it I realized that it was wrong for the fact that I didn't subtract the tangent point on the other side of the roller wheel (the 644 dimension) In other words the leading edge of the cam lifts the roller wheel. The sketch shows 90 degrees of rotation CCW. At that point the roller wheel is off the hump and back against the tangent point on the other side. As unconventional as it looks to the eye it is correct for 90 degrees of duration. I didn't draw this to any particular size so if you plug in your dimensions the hump will change a little but it will still look similar to what I have drawn. The picture attached show the cam for my tiny side shaft engine.
gbritnell

[Brian Rupnow]

George--This is what I had in mind. IF I can use my 1" diameter endmill to put the 1/2" radius in my cam, then the cam follower should follow it without being overly concerned about tangency points. An 0.100 rad at top of the ramp should in theory let my .866" diameter cam follower roll fully around the cam, remaining in contact with no "bump" when it reaches the ramp. Both the ramp on the rise side and on the fall side would be identical.--Brian

[gbritnell]

Brian,
The problem with your sketch is that the cam follower is drawn offset from the centeline of the cam. By moving the cam follower to center it changes the point where it starts to rise, and fall. If you take your top view and move the cam follower to center then carry that radius shape down to the front view that will give you an offset from the centerline. Now take your sketch and rotate it 90 degrees or lets say you have the exhaust open at 30 degrees BBDC and close at TDC then that would be 210 degrees crank duration (105 cam) so rotate the cam 105 degrees and draw the arc created by the end mill but you have to shift it to the other side of the center line. Now you get a shape like I had pictured in my sketch.
gbritnell
PS. Unless your cam follower is actually offset from the center line.

[Brian Rupnow]

So--Here we have it---a beautiful cylinder head. All the holes ended up in the right place. When I went to use my electronic angle finder to measure the 20 degree sloping face, the battery was dead. Trying to get the back open to put in a new battery broke the plastic housing. I have an old gravity actuated angle finder that I used instead. It's kind of critical, because if the angle is off by much, the spark-plug hole which is drilled blind can come out into the valve guide holes and that is very bad ju-ju. I was lucky, and the sparkplug hole come out on the far side exactly where it was supposed to. You will also see, setting beside the cylinder head, a pair of Viton O-rings which will seal the interface between the outside of the cylinder and the water reservoir.

[gbritnell]

Brian,
I copied your sketch and scaled it as best I could and made this drawing. The top view shows the .866 cam follower on the same centerline as the cam. This gives an offset of .334 from the centerline. Now I rotated that line 105 degrees and drew a second line .344 to the other side of center. This represents the tangent point of the cam follower to the cam.
This is what your cam should look like.
gbritnell

[Brian Rupnow]

Dang George--I guess if it was easy, everybody'd be doing it!!  From what my cad models are showing me, it isn't as simple as the cam follower rolling around the face of the cam. Everything would be fine if the path on the cam was a single line with no width to it. But--The path does have width to it. Which means that the cam follower is not only following the theoretical path---It is sliding on the surface of the face while it is travelling. This sliding action is going to demand that both cam and follower are going to have to be hardened. It is beginning to look like I should scrap the cam follower bearing and substitute a 1/2" diameter piece of hardened 01 material for the cam follower and the cam itself.