Press Fit

[Stephen Kowski]

Originally Posted by sanddrag

"A few thousandths" is FAR too much interference unless you have a really big bearing. Ideally, you're looking for something in the neighborhood of 4 ten-thousandths undersize for that size of bearing. You can get special undersize reamers to do press fit bores.

Let me ask, why do you need a press fit anyhow? In FRC, they're often more trouble than they're worth. Just make a good close slip fit and use part geometry to keep everything in place (flanges, shoulders, counterbores, snap rings, etc).

to elaborate on sanddrag's post for a press fit it is a pretty simple process....look at this chart: http://www.turnedoutright.com/wp-con...rill-sizes.jpg

You select a drill size two sizes smaller than your desired hole size, go one step up to ream out the hole, and then press it on

for example: if I want to press onto a 0.250" shaft, I would go down two steps to a size D drill (0.2460"), then ream out the hole one with one size larger (0.2480"), then press....and you are done

[George C]

for example: if I want to press onto a 0.250" shaft, I would go down two steps to a size D drill (0.2460"), then ream out the hole one with one size larger (0.2480"), then press....and you are done[/quote]

0.250" is the same as an E drill. D is one size down. Two sizes down would be a C = 0.242"

[Peck]

Originally Posted by George C

for example: if I want to press onto a 0.250" shaft, I would go down two steps to a size D drill (0.2460"), then ream out the hole one with one size larger (0.2480"), then press....and you are done

0.250" is the same as an E drill. D is one size down. Two sizes down would be a C = 0.242"

he isn't talking a generic drill set, he is talking a real machine shop's drill set where they have the drills accurate to tenths of a thou. An E drill is a generic and it is typically .250, accurate to the thou. If you are pressure fitting with generics... be really careful and take plenty of measurements.

[DELurker]

All of your bearing fit information (radial ball) can be found in the chart on page 131 at http://catalog.timken.com/WebProject...ookcode=eng11#, with the codes defined on pages 132-145.

If you can't make the holes to the specifications listed there, 0.001" undersized should work. I've found that Loctite Retaining Compound 641 works well in other applications, if you want to make the hole right at the bearing diameter.

As for press fits into thinner materials, I would recommend a flanged or snap-ring bearing with either retaining compound under the lip or 3 screws at 120-degree intervals.

[MrForbes]

fwiw, we made a couple gear boxes this year using flanged bearings and polycarbonate side plates, we discovered that drilling a hole with a Forstner bit will provide a sufficient press fit. There are lots of ways to do things like this.

[pmangels17]

Also note that polycarb cracks when you put loctite on it.

[AdamHeard]

This is an area where standardizing in design can make things a lot easier. If you only use a few common sizes of bearings, then you can get the appropriate reamers for these bearings (at some expense, but not awful; and due to design standardization they're used almost everywhere on the robot).

We have a press fit and slip fit (press being .0005 to .001 under depending on the availability of the reamers) for most bearings and bushings we use.

Most 973 robots over the past few years have most gearing running on plates that were waterjetted then reamed out by a reamer in a hand drill.

The reamer is really the best compromise between precision and the ability to use it on any given part and machine. It also allows "fixing" of damaged or out of tolerance parts.

Originally Posted by squirrel

fwiw, we made a couple gear boxes this year using flanged bearings and polycarbonate side plates, we discovered that drilling a hole with a Forstner bit will provide a sufficient press fit. There are lots of ways to do things like this.

Materials like polycarbonate that are much softer/weaker than the bearing itself will allow a much greater range of bore sizes without undue damage to or efficiency loss in the bearing.

[Cory]

Originally Posted by Peck

he isn't talking a generic drill set, he is talking a real machine shop's drill set where they have the drills accurate to tenths of a thou. An E drill is a generic and it is typically .250, accurate to the thou. If you are pressure fitting with generics... be really careful and take plenty of measurements.

There is no twist drill that can create a hole that is accurate within a tenth.

The diameter of the drill itself may be ground more precisely than on a cheapo drill, but the geometry of the tool inherently prevents it from making extremely round/accurate sized holes.

[Brandon Holley]

Originally Posted by Cory

The diameter of the drill itself may be ground more precisely than on a cheapo drill, but the geometry of the tool inherently prevents it from making extremely round/accurate sized holes.

In addition to this, any sort of runout on the drill/lathe/mill you are using will compound these issues.

Reamer is the way to go if you want to have repeatable results from plate to plate. Drilling holes, milling them, laser cutting them or waterjetting them will all result in a different sized hole for the same "as designed" size.

The best way to deal with it is to ream.


One understated method of holding bearings in is green loctite (gap filling). Having the bearing slip fit inside a hole, and using appropriate green loctite does a good enough job of holding bearings in.

Also- the ball bearing/ball peen trick is a nice one if youre in a bind. We have a nice 1.5" diameter steel ball bearing. If a hole is a little bit too large, we can just stuff that bearing into the hole and whack it with a hammer. It creates a little curl of material just big enough to bite the bearing and hold it in place.



-Brando

[AdamHeard]

Originally Posted by Brandon Holley

Also- the ball bearing/ball peen trick is a nice one if youre in a bind. We have a nice 1.5" diameter steel ball bearing. If a hole is a little bit too large, we can just stuff that bearing into the hole and whack it with a hammer. It creates a little curl of material just big enough to bite the bearing and hold it in place.
-Brando

On Emperor Swerve we had a 37mm bearing bore that we attempted to waterjet to exact size cutting super slow (like .7 in/min). The walls ended up being nontapered, but the kerf was larger than normal so the hole was easily .002 oversize.

We used some punches and ball peen hammer to rough up both the lip, and the inner diameter of the bore quite a bit in an even fashion around the diameter, and finished the assembly with bearing loctite.

We were able to "fill" that .002" gap in a really high load application with these methods and the bearings still have not loosened.

[Gray Adams]

A slightly undersized reamer is definitely the way to go. We've done all kinds of things in the past, but if you want a hole precise to .0001", use a good reamer.

We machined bearing blocks this year on our CNC mill, and some turned out as slip fits, some were press fits. Keep in mind this is the same end mill, same program, same bar of stock. You would think they all come out the same, but there is never a guarantee. Between a bit of high grit (400 or so) sand paper and snap rings, things were easily "good enough" for FRC. But...

Ideally, I would drill it undersize for the reamer and then ream it slightly undersize. Picking the size reamer you want is the hard part, but I would think .0005" under or so would be good enough. Other methods may work, but a reamer is going to give you the closest thing to a true cylinder that you can get.