Pottymill horizontal engine build

[Thor]

Hi Todd,
 
Your mill engine is coming along very nicely, keep up the good work and post more photos as you progress with the build.

Thor

[cfellows]

Very nice work indeed!  Thanks for the many, high quality pictures.

Chuck

[tinglett]

Well, it's been ages since I've checked in.  The main reason is that life and work caught up with me (so much for hiding in the shop!), and I also was heading into making a flywheel for this engine and took a little hiatus into thinking about the geometry of it all.   But meanwhile, I've been religiously following many builds including Don's grasshopper, Stan's down on the farm boy mini-series, and many many more.  I particularly followed Jo's battle with the 1802 bell crank engine which was quite inspiring to see her win against all odds.  I'll think more than twice about punting a project in the future.  And then there's crueby's kozo new shay build which I am certain is caught in some kind of time warp as nobody can work that fast in the universe I live in.

But you guys are certainly following those builds as well.  Back to this one.

The Flywheel

I'm starting the flywheel which is about 4 inches in diameter and is something I'd like to make fairly heavy to give this engine a chance of running slow even though it will suffer with my building tolerances.  A got a couple slices of 4 inch diameter cold rolled steel a few months back just for this purpose.  Hopefully I'll be using only one blank to get the job done!

I fired up Marv's flywheel app under DosBox on my android tablet and I got caught up in the calculations.  I got a snazzy new RT a couple months ago and it's barely been for a test drive.  This is my opportunity.  I'd like the work on the RT to be absolutely accurate, but I also know that if I screw up a little on the spokes, I should be ok anyway.  Thinking about how these angles might be calculated got me thinking about what the RT really does for a machinist.  It certainly implements the ability to work in polar coordinates vs. xy, but it also lets me rotate coordinate systems.  With that in mind, and with my (very) rusty programming skills, I whipped up a javascript-based web page that drew the flywheel with calculations.  Once I got the same numbers as Marv's I decided I'd better stop and just make the damn flywheel!  I'll get back to the software later.

So where to start?  It feels like I'm starting a new engine because I've been out of it so long.  I decided that since the flywheel was going to be so bulky, I'd better make a mandrel to hold it for finish turning.  I did that now as shown in the next few photos.  The shaft for the flywheel is 5/16 inch, and I plan to make the hub 5/8 inch, so I started with a 5/8 inch rod that was fairly beefy, turned it down, and threaded it.  I cut it off a little extra long in case I wanted to make a small mandrel on the other end for different diameter shaft.











So far that was pretty basic stuff, but I feel like I'm back into it again.  You guys have any advice for the mandrel?  I turned a short length near the shoulder so I can indicate off it when chucking the mandrel in my 4-jaw.  I did that while turning the rest so they would be true with each other.  Otherwise, I didn't put too much thought into it .

I'm planning to do the trepanning over on the RT to remove the bulk of the material for the spokes, so the flywheel shouldn't be turning with its full weight, but it still concerns me some.  I might just use it for tidying up the outer surface of the rim...but we'll see.

Finally, I chucked and indicated the flywheel blank.  The blank is pretty rough, and there wasn't really room for my DI, but I did pretty good.  It's hard to tell with the roughness, but I'm certainly within 10 thousands all around.  Maybe better.



And that's all for now.  It's too late to start turning it, so I decided to get a post out instead.  See ya soon.

Todd

[crueby]

Glad to see your build going again!

Well, it's been ages since I've checked in.  The main reason is that life and work caught up with me (so much for hiding in the shop!), and I also was heading into making a flywheel for this engine and took a little hiatus into thinking about the geometry of it all.  But meanwhile, I've been religiously following many builds including Don's grasshopper, Stan's down on the farm boy mini-series, and many many more.  I particularly followed Jo's battle with the 1802 bell crank engine which was quite inspiring to see her win against all odds.  I'll think more than twice about punting a project in the future.  And then there's crueby's kozo new shay build which I am certain is caught in some kind of time warp as nobody can work that fast in the universe I live in.

And we are definitely following your build too! Time warps are for cheaters - I prefer just stopping time when I go into the shop.... 

[b.lindsey]

Glad to see you back on the project also Todd. Nice start on the flywheel!

Bill

[Jo]

Hi Todd,

I particularly followed Jo's battle with the 1802 bell crank engine which was quite inspiring to see her win against all odds. 

I haven't won yet  the paint is too thick, the bearings won't go back where they should go. Did I mention the paint not sticking  After two weeks the crank is still not together but Kitti and Kiwi are so cute 

Sounds like you have that bit of CI nice and true. I normally only line rough castings with the edge of the lathe tool or in the case of a bit of CI round bar not even do that and just "bung it in a three jaw" 

Jo

[tinglett]

I made a tiny amount of progress on the flywheel today.

In the last installment I had just chucked and centered a nice beefy hunk of cold rolled 1018 steel when I ran out of time.  It's about 4 inches diameter and a bit over 1 inch thick.  Now I went about cleaning it up a bit before taking it over to the rotary table.

I faced it off and was rather impressed that it was as flat as it was from a bandsaw.  I took a couple light passes to get to this point.  I think it was only about 30 thousands to finished.


Next I worked on the outer surface.  This crust was incredibly tough and was made further difficult by the fact that I didn't have much room on the cross slide to get over there.  You can see I'm using a carbide cutter which I've learned to dislike.  I just can't get them to cut like the HSS tools I grind (even though my grinding skills are *very* lacking).   I honed it a bit, which helped, but it still was very poor.  Of course it was fighting the crust...but didn't do well after it finally got through either.  I played with height a little, but couldn't find an improvement.  But I got through it.


This was the best it would cut.


Here's the first side done.


Then off to the other side.  Indicating first.  You can see I'm challenged to setup a DI here.  I probably should have got the magnetic holder out, but I found this worked good enough.  I'm planning to do the final work on a mandrel so perfection is nice but, I don't think, required here.


It came out good.  Better aligned than I expected.  Ok, maybe my expectations were low...but that's what happened the last time I did this.


Finally I faced it off.


Next was a matter of drilling and reaming for the shaft.  I went through 3 drill sizes (not all shown here), ending with N, and then reamed to 0.311.  I measured the drive shaft at 0.308 so this was the best I could do.






I got a real nice fit.  Now maybe this isn't the best thing to do with the engine, but at least I can say the lathe was unplugged .  I slid it on about half way, not wanting to get it stuck.  That's a mighty massive flywheel at the moment.  I think it'll need a little thinning to be in proportion with the engine.  We'll see after I do some work on the RT.


And now to the RT!  This is getting exciting.  As I've mentioned before, I've never really used the RT before, other than a little playing around.  I'm not quite ready to trash this flywheel just yet (!), so I'll be experimenting for a while without cutters.  I need to align the RT, and align and clamp the flywheel first.  I've read and seen a couple videos on this, but any advice would be welcome.  I don't currently have any alignment gizmos, but the RT has a MT2 and I happen to have a MT2 center if that's useful here.  And of course my flywheel has a nice centered 5/16 hole I can easily use as a reference for some drill rod (I think I have some -- need to check for that diameter).

Any pointers to web sites or youtube videos would be welcome also.  Once I get it clamped and centered, I plan to hit it with layout dye and scratch out the lines using the RT and mill dials to verify I know what I'm doing, and that my measurements are right.   I have my handy sheet from my flywheel app printed ready to go, so you can see what I'm aiming for for the flywheel design.


Well, that's it for now.  I'll be in learning mode for the next few days, but I'll keep you posted.  Hopefully it won't be too amusing .

Todd

[rodw]

Todd, Looking great. I am about as experienced with a RT as you are. I have an MT2 taper a mate machined on a CNC lathe that has a straight shank I can grab in the spindle. I have used that to centre the table, a sweep with my Coax indicator always shows that it is fairly close but not perfect. On the table like your setup, when using the Coax indicator, I found I had to rotate the spindle by hand because the T-Slots were in the way. I think you could attach a magnetic based indicator holder to the spindle for final adjustments.

I've been working on a stepper based rotary controller driven by an Arduino that has become a bit of a monster due to the feature set. I'm not sure if you want to get distracted right now, but it is all here http://www.homemodelenginemachinist.com/showthread.php?t=24118

It accepts input of the angles to the nearest second and you just push left and right arrows to move a division either way.

One day, I'm going to make this engine with your fantastic build log to guide me!

[Hans]

Nice work, Todd! That was quite a large turning for that 7X lathe. Am I correct to assume that you are using a 5" chuck?

Hans

[tinglett]

RodW,

That's a cool project you've got going!  I've really been wanting to dive into some gizmos for my lathe/mill, and I want to use it as an excuse to program an arduino, but I've been a good boy and have resisted the temptation for now so I can get a few engines behind me first.  But it's cool you are moving that one along over there (though my eyes boggle with the myriad ads on the hmem site...pffft).

DROs are very high on my list.   All along, and this is probably very naive since I haven't programmed an arduino yet, I've been thinking of marrying an arduino with an Android tablet.  Android tablets and phones are insanely cheap for the hardware they provide, and the arduino can do all the controller/realtime stuff.  I wanted to do this for DROs and happily ran into yuristoys a year or so ago and see he's already doing it.  So I'll probably go that route.  But we'll see.

Meanwhile, I'm trying to do a little giveback here by posting beginner stuff (which I CAN do ) and will soon slide into a little software, I am fairly certain.  I should also look to see if there is a way to donate a little cash to the cause of running this site.  There's a lot of volunteer work here, sure, but there is also physical hardware, network and electricity to pay.

Hans,

Yes that's a 5" chuck.  In particular, it's the #2346 chuck and adapter from littlemachineshop.   When I bought it I was thinking I needed a 4-jaw and needing more capacity.  It is VERY big for a 7X lathe, and I think a bit overkill.   I did another engine last year with, if memory serves, a 3 or maybe 3.5 inch flywheel that didn't fit the standard 3" 3-jaw, but was small for the 5" so I had to move it in a step in the jaws.  That was tough on the nerves with those jaws flinging around at me.

I should have mentioned that chewing through the crust of this flywheel really did put a heavy strain on the lathe.  Those plastic gears in the head (the hi/lo gears) clicked a few times and I thought for sure I was going to trash it.

BTW, is there a real use for the high speed setting on the 7X lathe?  Honestly, I've never thought to shift it there before.

Todd

[tinglett]

My apologies for the excruciatingly slow progress.  Hopefully excitement will build over the weekend, but for now I centered up the RT.

I found a use for my MT2 center...it helped me get the RT roughly on center like this.  I had a 3/8 inch collet in the mill because my DTI will fit into it.  Hopefully this isn't trashing anything.



While it was held down by the mill, I shifted the XY slightly so I could bolt it down in place.   I inserted a DTI (a cheap one, unfortunately) and found I got it to within ~10 thousandths all around.   So a couple adjustments on X and Y and it's way better than half a thousandth, or whatever my DTI can really tell me, anyway.



Not exactly a riveting story, but it surprised me how close it got so I thought it might be worth showing.  I didn't really have time to work in the shop tonight, but sometimes you've just gotta sneak in there to do something even if it's small, right?

Todd

[rodw]

Todd thanks for the encouragement on the software front. Its taken me a couple of years to get to here and then just when I thought I had it finished, I smoked an expensive stepper driver as I was setting up for what I thought was the last test session.. 

There is a setting in HMEM's profile to turn the ads off so I don't see them anymore.

Glad you got your RT centred and finished it off with an indicator. I've got the same table. This morning, my local machine shop had a sale so I went along and I saw they had a chuck backing plate for it with an MT2 taper pressed into the centre so I'm adding that to my to do list. It is hard to centre the chuck after you centre the table and the last time I did it, it moved half way through the job. Fortunately, it wasn't critical on the job I was doing, just annoying. I'll tighten things up next time!

[kvom]

That brazed carbide turning tool is inferior in my view to one that uses inserts.  I think you'll like turning with carbide a lot more if you switch.

Engine coming along nicely.

[tinglett]

Marking out the flywheel (part 1)

Since I'm new to using a rotary table (RT) I thought it wise to go through all the motions once and see if it will produce the cuts I want.  This is somewhat artificial because scribing lines on layout dye won't be accounting for cutting diameters, but I'm confident the practice will be worth it.

I was originally planning to do the trepanning to create the webs on the mill, but this flywheel is big enough I've decided to do that the "traditional" way over on the lathe.  But for now, I'm going to this test drive on the flat surface of the side of the flywheel as it'll be easy to mark without hubs and rims getting in the way.  But this does mean I'll need to repeat parts of this again when I'm doing the real cuts.

Here's what it will look like.  The dimensions are in inches.  I've been tinkering with the proportions as I go along.


Before I can mark it out, I had better learn how to center and clamp it down.  I realized the T-slots on my RT were smaller than on my mill.  So in the process of looking up their size, I remembered other accessories came with the RT -- including a clamping kit.  So I was in business.

First I centered the flywheel using 5/16" CRS rod.  I didn't have drill rod, but there was zero play so I'm pretty confident it'll be ok.


For marking it, I need the clamps tight so it doesn't wiggle around, but I'm not actually milling so I tried to push the clamps as close to the edge as I can.  They worked pretty good.



Ok...onto marking it out.  But first I had to think about how I was going to scratch into layout dye.  In particular, how was I going to scratch out circles to represent holes I will drill later?   Here's what I came up with.

My outer holes in the web will be 0.188 (3/16) so I found a bit of scrap 3/16 rod that was already drilled.  I turned it from the inside until it was sharp around the rim.


Over on the mill I gave it a little test.   Note that the MILL IS UNPLUGGED for this entire process.  All I did was lower the mill to contact this little scrap with layout die, and then I gave it a twist of maybe 1/8th turn.  It measured plenty good.  That wasn't something I was going to draw by hand.



The inner holes in the web will be 0.375 (3/8 inch).  So I repeated the process.  This time I ended up drilling.  It didn't produce a crisp edge, so I tried a countersink and it cleaned it up nice.





The last thing to produce was something for scratching the straight lines.  Instead of chucking a scribe, I just made a point at the end of a 3/16 rod.  In hindsight, hardening it may not have been a bad idea.  When it got dull, though, I'd touch it up on the lathe.


I pulled the flywheel from the RT and painted it with layout dye.  While it was drying, I went through the process of re-centering the RT.  I wanted to make sure the last adjustments I made were in the "pushing" (clockwise) direction.  I wasn't certain if I had done this yesterday...and I had to wait for layout dye to dry anyway.  I also marked the X and Y with a grease pencil.  I learned that trick from That Lazy Machinist youtube videos.   He has loads of great advice over there!

Now I can return to zero from the clockwise direction at any time, and these blue marks will get me on the correct full turn of the dials every time.


This one on Y is a bit hard to see...but it's there on the base casting.


With everything clamped down and zeroed out it was time to make some marks.  I decided that the circle for the hub and rim were harmless enough.  I dialed out 1/2 of 0.625 in X for the hub radius.  This was somewhat entertaining and gave me a chance to look around where everything was, how far it moved in one crank, etc.  It was nice not having to deal with a cutter.



The rim was a little more tricky because the clamps were in the way.  Fortunately I'm only marking, so I simply raised the mill head as I got to them and skipped over that little bit of layout dye.



Next, I'll lay out radial lines to represent the spokes.  These are the dotted lines in the drawing.  I have 5 spokes so these are 360/5 = 72 degrees apart.  I turned my RT to zero.  This is a grizzly H7527 rotary table, but I suspect most look like this.  I had to ensure the scale up on the table was at zero as well as the handwheel and vernier.   Easy enough.  I decided to draw the lines down the X axis (to the right) since that area on the mill table is easy for me to see.  Any of the 4 directions would work, of course.

Here is my RT's zero setting.


I had already scratched out the first radial spoke so was able to immediately rotate to the next spoke at 72 degrees and scratch it out.



That looked fine.   So repeat 3 more times!  One at 144, one at 216, and the last at 288.







Now the dial was returned to zero (another 72 degrees) and everything lined up as expected.



One super duper simple RT task behind me.   Next is the fancier stuff.  I'll layout the hole positions by drawing the holes with the tools I made, I'll sweep out the outer arcs, and then I'll position and draw the cuts between the holes which (hopefully) will simply follow the software's calculations.

Todd

[tinglett]

Marking out the flywheel (part 2)

In part one I made some scribing tools and got the radial markings for the spokes made.  I also scribed rings representing the hub and the rim.  Now I mark out the rest, which is the important stuff because this is what I'll be milling later on.   But for now this is practice and I found a few things I was doing wrong with my RT along the way.  Better now than with a cutter.   FOR THIS ENTIRE POST THE MILL IS UNPLUGGED.

Remember, I said the mill was unplugged.

Laying out the holes comes first.   When I'm doing this for real I'll simply drill these out.  But I wanted to make sure all the math works and the lines match up.  As a bonus I'll get to see it drawn up full scale by the mill.

I'm making a 5 spoke flywheel and using a single larger hole to make the bottom edge of the opening between each spoke.  Therefore the spokes are 72 degrees apart.  The inner hole is centered at 36 degrees (halfway).  So I rotated the table to 36 degrees and dialed out the X axis to the radius of 0.625 inches to find the first hole's center position.  My RT rotates 4 degrees per rev of the handwheel, so this took 9 handwheel turns.  Easy enough.




Layout of the 2nd, 3rd, 4th, and 5th hole couldn't be easier.  All I had to do was rotate the RT the spoke angle of 72 degrees which is 18 handwheel turns.  When I was finished scribing the last hole, I had to rotate a final 36 degrees (9 handwheel turns) to get back to zero degrees.

This one was at 108 degrees.


180 degrees


252 degrees


324 degrees


All these handwheel turns is very similar to the XY handwheels.  It's easy to lose count.  Over the past few months I've learned to calculate up front how many full turns and partial turns to do BEFORE starting to move anything.  More important, I write it down in front of me...usually in the form A+B, where A is the number of full turns and B is the leftover.  That leftover is thousandths on the XY handwheels, and now on the RT it's going to be the last +3 degrees + min + sec.  I need to figure out what I need to do with that.  Maybe it's an A+B+C'D'' or similar.

I've also learned that when counting a large number of full turns (and 18 is getting large), that it helps not to look at the numbers on the dial.  I mentally think of a rough position where the turn is "done" but comes up short...so on the last turn I can take it from there to carefully position to the exact spot.  Then as I turn it fast I tend to watch my thumb that's on the handwheel and count it when it goes past that position.  It really helps.  Otherwise I get distracted trying to spot the zero as it flings by, and that leads me to pause as it goes around.  And that leads me to wonder if I just paused ahead or past the turn.  I'm curious what other tricks people have for this.  DROs eliminate the need, I suppose, but the RT may not have a DRO.  So welcome to the club if that's your situation .

The outer holes were fairly easy too, except on the first one I realized that the spokes are going to be rather thin with the plan I had.  The 7 degree angle would have resulted in a spoke that thins to 0.162 inch (less than 3/16).  I added a little code to my software to compute that, and settled on 8 degrees which produces a spoke that is 0.212 inch thin at the outer edge.  On the plus side, 8 is a nice multiple for my RT.  Another lesson learned.

These next two photos show the 7 degree angle.




Marking out the outer holes was fairly easy.  I could do them all in one pass, and with them all being multiples of 4 degrees (i.e. multiples of full handwheel turns) I could alternate 14 cranks and 4 cranks to lay them all out.  BUT, to save on the brain, I went around in two passes doing the "odd" holes first (so to speak) and then doing the "even" holes.  These are ALL 18 cranks apart.  Not much room for error.









The angles ended up being 8, 80, 152, 224, and 296, but mostly I was paying attention to counting cranks.

Next I had to shift to the "even" outside holes.  Rather than back up the RT, I moved forward instead.  Since I was over an "odd" hole right now, I knew I had to advance 14 full handwheel cranks to get to the right spot.   As an alternative, I could have moved back to zero, and then moved forward 72 - 8 = 64 degrees (16 cranks).











The angles ended up being 352, 64, 136, 208, and 280.

Now is the fun part.  How do you connect the outer edges of the inner and outer holes?   This is what Marv's flywheel program computes for you.  My little prototype software does the same and tells me I need to rotate the table 4.61 degrees and shift it by 0.220 inches.  For grins I ran Marv's flywheel and it produced an angle of 4.563 degrees and a shift of 0.219.  The discrepancy might be rounding within the math (I used some numbers 4 decimal places out).

First, the shift.   I had to back up my Y axis 220 thousandths.   My handwheel is 62.5 thousandths per full turn, so this is 3 turns minus 32.5 (going backward).  So I backed up 4 full turns and then advanced forward to 30 on the dial.  This represents the -32.5 (remember the dial is 62.5).  You can see the setting in this photo, and note my little grease pencil mark from before.  I am not going to mess with the dial so I can easily find zero again.  I find that trick VERY handy!



Now I had to dial in 4.61 degrees.  But since I pulled back the Y axis, I need this to be a negative 4.61 which is 360 - 4.61 = 355.39 degrees.  I'm treating the RT just like my other handwheels.  It has backlash so always go forward.  I could have backed it up extra far and then gone forward, but it goes round and round just fine (unlike X and Y!).

So what is 355.39?  I need to change it to degrees, minutes, seconds, so I'll make a mental note to fix my software to report that way.  But for now, that's 355 degrees.  That leaves 0.39 * 60 minutes = 23.4 minutes.  So 23 minutes and 0.4 * 60 seconds = 24 seconds.   355 degrees, 23 minutes, 24 seconds.   My RT only has a resolution of seconds to the nearest 20, so I'll be using 20 seconds.  So I set the angle, and realized later I didn't do this right!

I think you can see enough in the photo.  I have it set to 355 degrees, which is really 352 degrees via full turns of the handwheel, and 3 more degrees as a partial turn of the handwheel (note the 3 near the top right side of the photo).  Then I counted out 23 minutes...two major ticks for 20 minutes, and 3 minor ticks for the remaining 3.  But this was a mistake.  Later I realized there are 2 minutes per minor tick.  So I should have advanced it to one minor tick past 20, then shifted it forward so the +60 vernier tick lines up (the one at the bottom) and then advanced forward a bit more so the +20 vernier tick lines up.  It's easier than it sounds, but this exercise was great to force me to work that out.



BTW, my mistake set the angle at 355 degrees 26 minutes 20 seconds.  Converting this back, it's 355.44 degrees.  That's 0.05 degree off.  Over a length of 0.8 inches this will result in a Y error of 7 ten-thousandths.   I had to do the math to understand why my scribing didn't catch this. Whew.

Sorry...I imagine a lot of experienced RT users are sleeping right now .

In the next installment I'll scribe out these lines.

Todd