Effer, a spare parts story.

Discussion in 'CNC machining projects' started by BattyZ, Jun 21, 2021.

  1. BattyZ

    BattyZ Silver Banner Member

    Here is the CAD'd plan for the rotating base and the rest of the table to the linear rails:

    Screenshot 2021-12-26 152640.png
    Screenshot 2021-12-26 152554.png Screenshot 2021-12-26 152746.png


    -Not really my original idea at all. That idea would have required some LARGE castings and just am not set up for that yet. These are all smaller modular pieces that can be cast and machined in my VMC.
    -What is not shown is that the bed itself will be cast into 2 or 4 pieces and bolted together.
    -The three single-fastener legs should allow for shimming if necessary.
    -There is a taper roller bearing and a thrust washer that will allow for preload to be set for no play.
    -Just realized the one table mount is not constrained correctly.
    -The legs do clear the table mounts (pretty hard to see)
    -Provisions to securely attach the ballscrew nut need to be added. However, looking at the side profile there is ample room for a rotating bull nut assembly which kind has been the end goal so that might happen!

    Feel free to openly criticize this design with only 1 caveat; you have to have a better idea to share!
  2. Al2O3

    Al2O3 Administrator Staff Member

    Looks good to me. If your rail mounts for the (Y? axis) were just a little taller could you flip the (C axis?) servo 180 degrees from its current position so it didn't extend above the rotary table?

    You have so many Axis I can't keep track of which end is up....LoL!

  3. BattyZ

    BattyZ Silver Banner Member

    Looks like about 22 lbs and 3 ft long. Have also poured some 12X12x2 inch thick plates.

    uhh you might be on to something here... (BTW Kelly, you nailed all the axis names!)
    20211226_190740.jpg 20211226_190557.jpg
    The cad shows a distance just over 4 inches from the table to the underside of the bed. If I lifted the bed up about 2 inches more, I could fit the 130W servo underneath possibly. The other servo is 180W variant. (64 oz-in vs 85 0z-in in torque, which are both multiplied by 10 or 20 depending on setup...peak torque is 2.5 times the listed and can be held for 10 seconds at a time.) I would most likely end up losing 3 inches of travel said and done but would gain rigidity and having the stepper out of the servo hidden.

    Got the misfit-of-a-casting to stiffen up the Z-axis installed. At nearly full extension of the Z, and with my shoulder on the frame whilst arm-barring the back of the plate, I was able to push it out by about 40 thou.
    20211226_184346.jpg 20211226_184334.jpg 20211226_184411.jpg 20211226_183446.jpg

    40 thou is a bit of play still although it is loads better than what I was machining with before. However, if the axis is choked up 4-5 inches, the play quickly sinks below 10 thou and that is most acceptable. The tone of the machine with the spindle on is much more 'dead' now. Effer also looks a bit more serious too.

    Last item: SKEW

    For the longest time, I have been depending on skew correction to correct non-perpendicularity between X&Y.

    Actually took the time to take some real-world measurements and then some high school geometry to realize I was actually way off physically. Without boring everyone to death...I took all but one fastener loose for the gantry. This way I could rotate the whole gantry about that point. Did just this, clamped the gantry in place and ran a skew test. Skew test being 4 dimples then measure the diagonal distances for squareness. Did this multiple times.

    Resized95Snapchat-157827015395603585930727196.jpg 20211226_204640.jpg 20211226_204628.jpg

    Effer is as square as I can measure for now! Told ya I was off by a mile. This guess n check method was time-consuming, glad it is done. Skew factor: 0.0. Time to machine up some more upgrades.

    Attached Files:

  4. Al2O3

    Al2O3 Administrator Staff Member

    As a rule of thumb, Morgan Salamander crucibles weigh about 60% the rated brimful aluminum capacity.

    Salamder Super Cruble Weights.jpg

    So and A20 about 32lbs plus shank, which weighs about 15lbs. The shank is a bit overdone because it was made for a pouring cart but all up that means it's approaching 50lbs. In my younger days I wouldn't bat an eye at slinging twice that but now that I've uhmmm matured, I (and my back) have draw the line there as far as free hand pours, but must say, at a 4ft pour height and 30 second pour, I added a steady rest to my pouring rig.

    The A60 full with shank is 135lbs......enter the gantry crane and hoist.

    Gantry Crane Build

    30 Gantry Assy.JPG

    For your part, it's more the height than the weight, but that's very similar to my intake manifold with gating and in the flask it ends up about 4ft+ to the mouth of the pouring cup. Here's my extended flask. It also gets the extra large economy size pouring basin......aka, the big gulp.

    Pouring Basin

    38 Flask Extension.JPG

    Was that lost foam? I find casting chunky piece more challenging than thin walled pieces. Gating and feed system are critical.

    Is that any different than tramming? Looks like one in the same.

    Now that I've sufficiently plugged up your thread, I can finish my coffee....:)

  5. BattyZ

    BattyZ Silver Banner Member

    Appreciate the links Kelly, helps me and any passerby find the good stuff!

    Those plates were open face sodium silicate cast. Blanks per my semi-permanent die process.

    Is that all you got?! lol Just took a gander on the Jets website for the Morgan Salamanders. The A25 currently piques my interest for a good size.

    After yesterday's work went to re-machine the B tilt axis part. Was not happy with the part-flipping alignment. This time the part flip only went slightly better. Ditched the bottom fence. Bolted 3/4 MDF to the table, faced it a couple of times and put a bunch of holes in it.


    Holes are quarter-inch, bored with a 1/8 in endmill, and on a 2 by 2 inch spaced grid. Found some snapped endmills to be dowel pins and...

    20211227_145528.jpg 20211227_151537.jpg

    Best flipped-part-fitment yet! I also dug in 30 thou on the tooth profile. Decided to push the envelope a little more.


    This yellow-green coating is the non-ferrous variant of SuspendaSlurry. Going to try 2 coats and cast it. I know it is not the intent of this investment slurry. It is supposed to be fired completely before the cast. I think some of us on here will be curious to see how it turns out.

    Kelly, you should be a redistributor for your commercial coating for LF....just sayin'.

    Oh, in betwixt hee-hawin' around on the CNC router trying to flip parts I made progress on the metal kiln cart. Probably a thread coming on that.

  6. BattyZ

    BattyZ Silver Banner Member

    Finally got some casting done.
    20220114_134035.jpg 20220114_130952.jpg 20220114_130930.jpg
    The short feed I think I had the metal too cold. Need to get better at keeping thermocouple batteries around.

    The last one was excellent. Then realized I don't have a lot of sand or boxes to cast with right now. Need to invest in some kegs or metal barrels.

    So the suspenda-slurry-lost-foam casting, one of the more violent pours gas wise. Which makes sense, gas was trapped and the only way out was up. Showed some excellent detail. I'd imagine the strength of the casting ain't at all what it should be due to bi-films caused by bubbles and what-not. Maybe 2 coats was too many? Didn't seem to get full coverage with the first? If I left bare foam spots for gas to escape eould this work better?

    All of these molds were dipped. Hence the very low amount of inclusions. Dipped in the drywall mix Kelly and I discussed earlier in tbis thread.

    The next parts were brushed on for the coating. I expect not as good results. I will start doing a better job machining the feeding system into the foam. CNC doesn't care how long it takes anyways. Lol I did like batch window maching five parts at once though.
  7. Al2O3

    Al2O3 Administrator Staff Member

    Did the suspendslurry easily release from the casting? I've never been motivated enough to try it in LF..........substituting a very expensive coating that may not work well for a very inexpensive coating that does......but glad you gave it a go ;). I've also never tried fully shelling a foam pattern and burning it out before pouring because it takes so much more effort and gains little to nothing. Why do I say so? Why span days dipping/drying/sanding multiple coats when one dip will do? The one dip will produce the same detail at least as far a the finish that is possible on foam patterns goes? The pattern will become so heavy the foam wont support the shell. Then you have to fire it? Then for all your effort you get to chip the shell off with a hammer, chisel, and media blast.

    I much prefer, dip it, vibe it in loose sand, pour it, dump it on the ground, quench the casting and blow off any residual coating with compressed air......done. With a coated pattern in hand my last three sessions have been <1 hour from cold start to casting in hand, cleaned up, and put away.

    I think either, if thinned to the proper consistency should give equivalent results. Folks tend to cake on brush coating too thick, and dipping with thinned mud can require two dips, but it's fast and produces uniform and consistent results. Always said, if it turns the pink pattern white, it's thick enough, and thin means more permeable. A drying box can be helpful.......<105F. I close the door to the bathroom in my shop and the heat vent in there makes for an excellent drying room in the Winter time. So does a cardboard box place over a floor vent in the house.

    I've taken to machining in at least the attachment points of the feed system on the pattern, and sometimes more if it doesn't make the pattern size grow too large.

  8. BattyZ

    BattyZ Silver Banner Member

    To add some more detail: I dipped this twice in the slurry. A day apart when I was at the shop. Days later got around to melting metal so vib'd it in some sand and poured.
    When I went to quench the casting made some crazy sounds, but the black (partially fired??) just flaked right off.

    I was curious! Suspenda-Slurry is Sus-spensive but if one dip coat will give great detail might be worthwhile? Especially since it is available in 5 gal buckets.

    Was about to quote and reply the rest of your message but that would just be clogging up the server to say I agree lol I do think I need to thin my coating some more and get to a point where I can pour a cup of the compound over said parts and get good results. That would be a great way for us home gamers to get the best mileage I think.
  9. Al2O3

    Al2O3 Administrator Staff Member

    Sometimes if I have a fragile pattern that I think will be broken by buoyancy forces, I ladle the coating on while hanging over the vat......it's fair game. I have also plugged internal passages to both exclude them from coating or to slush mold/coat only an interior core. Some patterns become very heavy when initially coated and there is risk to deforming or breaking the pattern (usually sprue). That's why I stick the sprue on after coating.

  10. BattyZ

    BattyZ Silver Banner Member

    20220118_132019.jpg 20220119_153232.jpg 20220119_153148.jpg 20220119_170553.jpg
    Things were coming along smoothly! ...That is until I received a 36 part side job.

    The 2 sides of the t-slot bed machined up without a hitch. Coating and casting may be a different story. Will have to be clever with the feeding system to keep everything straight. Also contemplated gluing 2 steel flat bars to keep things flat. Using the flat bars as pseudo "chill" plates but primarily as a substrate.

    All 7 legs turned out great and useable as is. Some of the other parts will need a quick touch up with the big CNC.

    Diablo Blades makes a 40 some tooth non-ferrous blade for skil-saws that really is the cat's meow for degating. Although it is a kevlar glove and face mask kind of affair.

    The three-phase furnace is doing a great job. Just flip the switch and go.

    Attached Files:

  11. Al2O3

    Al2O3 Administrator Staff Member

    Another episode of tools making tools! Nice haul. You going to heat treat those (assuming they are 300 series). A T5 is easy and gains a lot of strength.

    Those are going to be a challenge. What are the approximate dimensions on those and what will they weigh? And the T-slot dimensions? They will be the biggest part of the challenge. I think you'll have a hard time without them being vertical in the mold.

    Not sure of the T-slot clearance or machining allowance that may have been allowed for them and the surface but the chances of having a little flaw or bubble booger along the lengths of the slots is pretty good, and enough to ruin your day fro sliding T-nuts. I'd dip them in soapy water then dipped them multiple times in thin slurry. Even so, it may be difficult to keep them flat (enough) to be a machine bed. As cast, it's pretty easy to push the castings around if you have a rigid flat surface.

    Have a look at the bottom of post #1


    I don't do them that way anymore. Many times I'll do two back to back and that makes them pretty stiff, but I've found it pretty easy to straighten them if/as needed so will do one at a time. Like here:


    If you have the pouring capacity, gating them together back to back may be a good strategy. Apply the gating while being held flat. Fill the mold with sand at even heights on each side or the weight and uneven sand pressure will deform the patterns.

    I use a similar blade. Also helps to apply a little wax/grinders grease here and there.......cuts like butter.

  12. BattyZ

    BattyZ Silver Banner Member

    Yes to the heat treat. The halves are 12x33x1.25. I plan to cast them 30 degrees away from vertical. A long time ago I fully caked one side of an LF sign mold and left the other sides uncoated. Worked very well. Was thinking about trying this with the T-slot side. There will also be the center rotating piece which will be 15 inches in diameter and 1.5 inches thick.

    That MDF substrate is a trick idea! Might be implementing that as well. Thanks for sharing that.

    I do believe I will have to do some cleanup passes on the T-slots themselves once installed.

    This would be ideal but I have neither the crucible of lift-off furnace for this....yet. I can also break this halves into quarters and redo the cad so they are thicker. I may be happier with this approach. Still want to see how these turn out. Can always cut them up...speaking of,

    Water Displacer Formula 40 does the trick for me but I should find some of that wax to try.
  13. BattyZ

    BattyZ Silver Banner Member

    Well, it has almost been a month. Machining jobs came in and ole Effer got put on the back burner.

    Using some advice from this forum I have got the bed pieces ready to pour.
    Resized952022021295161429.jpg Resized952022021295161435.jpg

    ....Once the coating dries. The Suspenda Slurry is only on the detailed side. I poured the slurry down in multiple ways to coat the undercuts too. Not expecting perfection. With the underside uncoated, gas can escape out the back, well hopefully. Using an MDF board as a backer. Need to make a custom box for casting then get to it! Did 2 coats.

    I am well aware that there is a 70% chance this will not work or will not get a usable casting. When that happens I will split the design into quarters.
    Tobho Mott likes this.
  14. BattyZ

    BattyZ Silver Banner Member

    Just like that, PLOT TWIST!

    This is what happens when I am left unsupervised! New old stock, for the B axis. My belt reduction design may have worked. This will work. Struck a killer deal. $1800 new, paid under 10% of that. Gotta cherish the obscure industrial corner of fleabay. Don't you love how people are so inclined to tell the MSRP price that no one would ever pay?

    93% efficiency at a nominal output torque of 22.2 ft-lbs.
    Servo motor can supply a torque of 85 oz-in X 50(ratio of the gearbox) is 22.1 lb-ft.
    Speeds are matched pretty nicely.
    Torsional stiffness is at 22 in-lb/arc-min.

    10-12 arc-min maximum backlash, did the triangle math. If the tip of the endmill is under 7 inches from the centerline (super long endmill/spindle combo), max runout(induced by gearbox) would be under 24 thou. For foam and wood that seems acceptable. An arc-min is 1/60th of a degree by the way.

    Depending on how it is mounted to the Z-axis I may lose some travel, maybe not. The tight(IP65 rated), enclosed motor/gearbox package is worth it. Also, it all bolts together. Have to make one mount piece. This could be done from flat bar stock on my big CNC or cast.

    2 other "twists" are in the works.
  15. BattyZ

    BattyZ Silver Banner Member

    Plot twist 2:
    right angle gearhead.png
    A big ole Parker Bayside RS115 precision right angle 100:1 ratio gearhead. It will be almost too big for under my table but for the right price I will make it work. The large output shaft should allow me to mount the C axis rotary table right to it making things a lot easier.

    Finally, hopefully, the last big thing I will need:
    AMB 1050 FMEU.png

    New "semi-automatic" tool change spindle. I went with the manual speed control for simplicity. Here is what I mean by semi-automatic:

    The lock rings use standard ER16 collets so you can switch between different diameter tools real quick. They have the clamping mechanism design against a hard stop so tools of the same diameter can use the same lock ring/collet and will always have the same offset once set initially. The other nice thing is there is a straight 4 hole mounting pattern for this. Should play really nicely with the B axis rotary. They say it is a lot quieter than a router but not holding my breath and it really doesn't matter. My compressor is still louder and the sound system is louder than that.

    Here is a cad look at the latest setup.
    latest setup1.png latest setup2.png

    This is just a quick mockup. Don't want to commit design time till I have the bad cast and ready to go.
  16. BattyZ

    BattyZ Silver Banner Member

    By "bad" I meant bed.
    20220221_120229.jpg 20220221_152315.jpg 20220221_132749.jpg Snapchat-1105616940.jpg
    2 pretty good castings! The casting box and filling method were a little unorthodox but I got good results. As for flatness, I need to degate them and see what we have. Took a very full crucible. Actually didn't have enough metal on both pours but having the feeding system on top, it gave wave first.

    Attached Files:

  17. BattyZ

    BattyZ Silver Banner Member

    Hopefully all the hardware I will need showed up. Everything feels to be in great condition. The right angle gearhead is a biggen. 20220226_122305.jpg
    The table and Y axis will have to be redone to suit. Not too upset because there are some nicely sized thrust and radial bearings for the output shaft. So will be able to directly mount the rotary table to that.

    Went from basically 2 tools to a whole arsenal. Tool swaps mechanically, are real easy. Just need to figure out the tool offsets and how to get them to automatically update after a tool change. Might be switching back over to Dynamotion from here.
  18. Al2O3

    Al2O3 Administrator Staff Member

    What size/style collet is in that spindle?

  19. BattyZ

    BattyZ Silver Banner Member

    ER16 Collet, allows shafts up to 3/8 inch. Once tool offsets are sorted and proven I bet my router collection will start looking more like yours!
  20. BattyZ

    BattyZ Silver Banner Member

    I have picked up some good side work here recently so work on this project has slowed almost to a stop. I might have some castings on the horizon so I try to keep on it when I can.
    20220309_200149.jpg 20220309_201936.jpg 20220310_192053.jpg
    Here is the B axis spindle mount. Allows the use of 2 tapered bearings to deal with cutting forces on the spindle. This was about at .0015 hydraulic press fit, tig'd in multiple spots top and bottom before the full run around. The set screw was the original plan for holding the shaft but now going to machine the end of a Lovejoy coupling right into the end of the shaft instead.

    I miss-machined the tolerance for the pockets that hold these bearings' races to the Z-axis. There is about .020 slop on the back bearing and 10 thou on the front. It will be a redo.

    The Z-axis plate redo, the coupler end machining, and a gearbox-to-Z-axis part are what's left for the B axis. Then place/wire limit switches and revert the control software back over to Dynomotion's KFlop. From here, Effer should be able to machine with 4 axis.
    Tops likes this.

Share This Page