Waterjet accuracy?

[asid61]

Generally I've seen waterjetting quoted as having an accuracy of around +/-0.003", maybe 0.001" as a minimum, and that there is a significant taper of a few thousandths generated on the part (which can be reduced by taking more time). However, I have heard that teams make gearboxes from waterjetted plates, so I wan't exactly sure how accurate it really is.
Can somebody from a team who uses waterjetting comment on its accuracy and precision? I would prefer actual numbers, but pretty much anything would be good, as FRC use is probably different from other things.

[Chris is me]

I can't actually answer your question at the moment, though maybe if I'm nice enough I can run some waterjet parts over to RPI's DMM and find out...

However, I can attest to gearboxes working with waterjet plates. The reason for this is pretty simple. It's one of those things that you can get away with in FRC due to the extremely short run time of an FRC robot. While in the real world, you wouldn't want to run a gearbox for 10,000 hours with +/-.003 tolerances, it's usually fine to run waterjet gearboxes for FRC purposes.

The taper (kerf?) is a real phenomena, but it's usually slight enough to ignore. For bearing holes, one should ream out everything they make anyway if they want a true precise press fit, so the taper isn't a big deal.

Our waterjet parts (relatively few each year) are reasonably accurate. Hole centers are located properly, if not the sizes themselves. I'd be comfortable doing just about anything 2D used on an FRC robot on a waterjet.

[asid61]

Originally Posted by Chris is me

I can't actually answer your question at the moment, though maybe if I'm nice enough I can run some waterjet parts over to RPI's DMM and find out...

However, I can attest to gearboxes working with waterjet plates. The reason for this is pretty simple. It's one of those things that you can get away with in FRC due to the extremely short run time of an FRC robot. While in the real world, you wouldn't want to run a gearbox for 10,000 hours with +/-.003 tolerances, it's usually fine to run waterjet gearboxes for FRC purposes.

The taper (kerf?) is a real phenomena, but it's usually slight enough to ignore. For bearing holes, one should ream out everything they make anyway if they want a true precise press fit, so the taper isn't a big deal.

Our waterjet parts (relatively few each year) are reasonably accurate. Hole centers are located properly, if not the sizes themselves. I'd be comfortable doing just about anything 2D used on an FRC robot on a waterjet.

Ok. I usually design with +0.000 distance between centers on my gears. How much do you account for?
EDIT: For reamers, do you use a 1.123" or 1.124" reamer? Where do you get that? Have you ever tried waterjetting, then finishing on a mill?

[InFlight]

If you look at the detailed dimensions of a AM or Vex gear box, they typically run + 0.003" to .0005" large between spur gears. In terms of reaming, we use a 1.125 as that's what's available.

This year we waterjet cut the entire structure of our robot. Accuracy would depend on the actual machine and it mechanical construction. It would make excellent gearbox plates.

[BrendanB]

Originally Posted by asid61

Ok. I usually design with +0.000 distance between centers on my gears. How much do you account for?
EDIT: For reamers, do you use a 1.123" or 1.124" reamer? Where do you get that? Have you ever tried waterjetting, then finishing on a mill?

For our bearing holes this year which were pretty much entirely Vexpro .5" round and .5" hex (1.125") we bought this one from Mcmaster #3107A33. It was a little pricey in comparison to the set we bought below but figured if it was a tool we would be using a lot of over the years it was worth investing in. The few parts we outsourced this year to a waterjet sponsor and another with a CNC we had bearing holes machined undersized and then reamed them out on our mill after some QC issues last year. It is slightly adjustable so we typically did our parts in batches separated by round or hex since we found the Vexpro round bearings required a little more to be reamed out.

We also bought a small set off of Harbor Freight for other sizes that would come up. http://www.harborfreight.com/11-piec...set-38577.html

This was our first year getting waterjet parts for a custom two stage gearbox and we were very happy with the quality. Never hesitate to reach out to the company/operator to discuss the parts you are looking to make and their recommendations on modifying your parts.

[Bennett548]

548 has quite a bit of waterjet experience, so I'll let you know how we do gearboxes. This only applies to our waterjet and waterjet software, so make sure you try a test piece before cutting too many parts.

We waterjet to 1.125" exactly. This makes the hole larger than 1.125" on the top, and smaller than 1.125" on the bottom. There is usually a burr where the waterjet starts and stops, so first we file that away with a half round file. Then we use a half round file going around the small side of the hole to evenly remove the far side of the taper, hopefully keeping the hole pretty round. You just keep test fitting the bearing to try to get a snug fit. If you go too far use some CA glue to hold the bearing in.

I know this isn't the correct way to do it, but it works well enough for us and hasn't really caused problems, so we will probably keep doing it that way. We have used an adjustable reamer too, but I don't think it made much of a difference.

Having a perfect hole for the bearing really helps it run longer and smoother, but FRC robots don't run long enough for it to matter.

We add 0.002" between small gears (like a motor pinion), and 0.003" between larger gears on an open gearbox. 0.002" for all gears on a closed gearbox. The extra distance in the open gearboxes helps when they get filled with dust and carpet fibers.

Edit: This is for 0.25" aluminum gearbox plates. As s_forbes's post reminded me, thinner stock means more of the hole is the correct size, reducing the need for filing. At 0.500" thickness we waterjet undersize and drill out the hole.

For all holes that will be tapped, waterjet undersize and drill before tapping. For example, for a #10-32 in a 0.25" aluminum plate, we waterjet 0.120" dia pilots.

[s_forbes]

For the last few years, we've done exact sizing for the bearing Outer Diameter (OD) and just deburred the edges of the holes. For a 1.125" OD bearing, we model exactly 1.125" in CAD and the cut parts come out as a very light press fit for the bearings. We press them in using a socket and a vise. The OD of the bearings we've accumulated over the years vary by about half a thousandth; the large ones need a vice to press in and the smaller ones can be forced in by hand..

0.003" is added between each set of meshing gears to make up for gear tolerances, and we haven't had any gearbox related issues so far. All of our gearbox plates have been 0.125" 6061 aluminum.

For those in Phoenix, we've been getting help from MarZee and have had fantastic results. http://www.marzee.com/

[RoboChair]

I wouldn't call this a normal use case, but... this was done via waterjet.

[protoserge]

Knowing the machine that is doing the work is key. You will want to interface with the machine shop to determine what tolerance the machine holds.

For FRC purposes, the accuracy of the average waterjet is plenty good. Remember, you can make a gearbox using a hacksaw and a drill press and it will work just fine. Of course it will depend on how well you perform the gearbox layout to make up for runout in the various tools (do you add slots for adjustments and gear mesh), the layout of critical dimensions, measure, and set up the workholding.