shrink-fit guards

NDallyn

NDallyn

Joined
Jun 22, 2006
Messages
814
So I had an idea:D .

I've tried both soldered guards and the press-fit method with JB-weld for hidden tang knives and have been dissapointed with both.

At work we shrink hubs onto the shafts of gas compressors by heating the hub to 350 F and sliding it on the shaft. The hub is machined to a smaller dimension than the shaft (we're talking somewhere between 0.002-0.0075"), when heated, the hub expands which allows it to slip over the shaft. When the hub cools, it shrinks down to the shaft dimension creating a strong, air-tight joint. This hub is then joined to a 12 cylinder Waukesha or Cat engine to drive the 2-4 stage compressor. No keyway, no welding, no bolts.

Could this principle be applied to guards on hidden tang knives? If I were to machine the tang to a certain size and then machine the guard slot slightly undersize, would it create a strong, atmosphere-free joint?

All you machinists and engineers out there let me know what you think.

Nathan
 
The only problem with interference fits is getting the part into place before it cools too much to get it there.
Fixtures to make the part get to where you want it in the shortest amount of time will help.
It will work very well if you can get the dimensions as accurate as you need. It will become a permanent joint.
 
We used to do the reverse at work and use liquid nitrogen to freeze the shaft and shrink it. There were times when things didn't get seated correctly, and that was that. Even our twelve ton press couldn't always remove the part!

I think your plan could work great if you nail all of your tolerances.
 
One other problem that I can see is that you cannot test-fit the guard (once it's on, it's on for good).
 
This post should help you nail the tolerances, so that you get it right each and every time.

The following chart is the thermal expansion rates of different metals per 100 degrees F. change in temperature per inch of linear length. Since these numbers are linear, you have to take into account both the height and width of the slot to achieve a tight fit. For example, take a 3/8" x 1/8" tang and slot. Since you have two surfaces of each length you have:

(3/8" * 2) + (1/8" * 2) = 1" of total lineal circumference of the tang and slot.

If you're going to use Yellow Brass with an linear expansion rate of .0113" per inch of linear length. You'll get .0014" of expansion per 100 degrees F. raise in temperature of the guard along it's 1/8" width. You could freeze the tang down enough to get a 200 degree difference, but the tang will only shrink by about .001" along the 1/8" width. This will give you a difference of .0024".

You also need to consider the height of the tang also. The growth of the guard and the shrinkage of the steel will be greater do to the increased length. You'll get three times the difference, so the change will be .0072". With a 200 degree difference. It's important to take into account your starting ambient temperature of each material. If it's 70 degrees in your shop, calculate the change from that point for each material.

The important thing to remember when shrink fitting parts is to make sure that your part is at the correct temperature all the way through. Don't just heat it up to the correct temperature and immediately put it on. Let it soak a few minutes. This will help you some with the issue that Howie mentioned.

It's always best to use temperatures that create what I call a #10 throw fit. You want to open it up as much as you can so that you could stand 10 feet away and throw it on! :D :D :D

This method will work, but requires close tolerances and quick action once the parts are at the correct temperature. Remember, you can freeze the steel, as well as heat up the guard.

As you can probably tell, I shrink fit a lot of parts on as part of my job.

Good Luck!

This chart list the linear expansion rate per 100 degrees F. change in temperature.
  • Adrmiralty Brass .0012"
  • Aluminum .0131"
  • Aluminum Bronze .009"
  • Cast Iron (Gray) .0058"
  • Copper .0098"
  • Cupronickel .009"
  • Iron, nodular pearlitic .0065"
  • Iron, pure .0068"
  • Manganese Bronze .0118"
  • Nickel Wrought .0074"
  • Red Brass .0104"
  • Platinum .005"
  • Silver .011"
  • Steel .0078"
  • Yellow Brass .0113"

Ickie
 
stevomiller said:
We used to do the reverse at work and use liquid nitrogen to freeze the shaft and shrink it. There were times when things didn't get seated correctly, and that was that. Even our twelve ton press couldn't always remove the part!

I think your plan could work great if you nail all of your tolerances.
Click to expand...

HAHAHA! I've seen two guys seize the hubs halfway onto the shafts (not enough heat, didn't have everything lined up before they started heating *millwrights*:D ). Then they have to come crawling back to the welders to gouge the wrecked hub off:D .
 
One other thing that I forgot to mention. Physics causes the guard to shrink away from the shoulder as it cools. This is even more pronounced on guards that are narrower in cross section at the handle material side. Because the cross section at the handle material is smaller it cools there first and becomes tight fit out at the end near the handle material joint. If the guard is made of the yellow brass above and is 1" from ricasso to handle material, you'll shrink away from the shoulder .0113". You'll have a gap this wide between you ricasso shoulder and guard and won't be happy!

If you make a jig and clamp the guard against the ricasso during the cooling down process, it will keep it from shrinking away. Dunking it in water won't stop it from pulling away, it will only cause it to pull away more quickly.

Good luck.

Ickie
 
NDallyn said:
HAHAHA! I've seen two guys seize the hubs halfway onto the shafts (not enough heat, didn't have everything lined up before they started heating *millwrights*:D ). Then they have to come crawling back to the welders to gouge the wrecked hub off:D .
Click to expand...

I haven't had a part stuck short of the shoulder for 10 years. When I was a millwright I got em' stuck all the time. A little knowledge can actually help.

Ickie
 
I think its a great idea, so long as your quick enough. What about using a copper pipe section hammered flat but not closed, so this can be slid over the blade after putting on the guard, and then used to regularly tap down the guard while it is cooling.. I dont think this would completely eliminate the shrinking away from the shoulder, but it should reduce it significantly I imagine? Good luck
 
The biggest problem is comparing a near perfectly round shaft and a milled hub to a multi surfaced tang and guard slot. The tolerances available for turned objects are immensely greater than for ground and filed parts. The fit would have to be perfect, and since the guard would be undersized, you couldn't test fit it. You would have one shot at getting it perfect, which would require VERY close tolerances. Since we are generally using softer guard metal than blade steel, a press fit should accomplish the same thing easier. In a press fit, the guard makes the final forming of the slot as it is forced on the shoulder. Same result, different physics.
Stacy
 
Thanks to everyone for their input. Stacy, I've tried to do a couple press-fit guards and have found that the softer guard material, nickel silver in this case, tends to distort around the joint. Maybe I'm not doing things properly.

One quick thought - Would milling out the guard slot to fit over the ricasso, in front of the shoulders on the tang, eliminate or reduce the issue with shrinkage?

Thanks again,

Nathan
 
striper28 said:
What would worry me the most is cleanup of the guard afterwards. I hate wrapping sandpaper around razor blades to clean up guard faces.
Click to expand...

What I have done in the past is to tape off the ricasso on the blade with painters tape and use a felt polishing wheel and some bufing compound on a dremel tool to shine up the guard face. Of course this will only work if you mirror finish the guard prior to mounting it and will not work on damascus guards. On most materials this should remove the surface oxides caused by heating.
 
You must log in or register to reply here.
Back
Top