How should I go about loosening knurled knob on centrifuge that's been overtightened?

[winchman]

It's about 2" in diameter and down in the chamber.



Unlike the picture, the rotor has a flat top right under the knob, so I can't get to it with soft-jaw pliers. Anything else I've got will scar the knurls. Any suggestions?

[lynnl]

I assume it's intended to be hand tightened/loosened, so shouldn't require much clamping force. Try wrapping with a strip of thick rubber or leather then grip onto that.

[_Paul_]

Have you got room for a strap wrench perhaps using leather for the strap to stop marring? or if you can find one to suit what about an oil filter strap wrench,



Paul

[kitno455]

Drive thin walled brass tube down onto it?

allan

[Noitoen]

Drill 2 little holes and use a pin spanner.

[A.K. Boomer]

I know im stating the obvious here but it's a centrifuge so verify if the threads are left or right hand - hopefully they have enough hanging out to see or have a mark or warning label.

then question if the machine can handle the static torque your about to apply without damaging it, or if there's a way of countering said torque by holding something lower than the knob,

Paul's got a pretty good suggestion there with the strap wrench and keep in mind when using them to favor the ratchet handle to the outside of the drive shank as this will keep your torque ratio up instead of having some of it go to waste and be counterproductive...

[Willy]

I'd say something along the lines of what Paul posted in post #3.
Perhaps a thick piece of leather on the knurls and a chain grip wrench, or the leather in conjunction with a strap wrench.
Leather the thickness of a good belt should suffice. As Lynnl mentioned being knurled it shouldn't take much to loosen it as it's only been tightened using the knurls and a good grip.
Kind of like a screw on lens filter, seems almost impossibly tight until you use the right tool to unscrew it and then it almost falls off by itself.

[EddyCurr]

Is the threaded portion of the knob internal (a nut) or external.

If it is a nut, what about adding heat to increase the thread ID - or would the longitudinal expansion increase fastener stretch/retension, becoming counter-productive? If heat could be beneficial, a non-marring application method is by conduction - heat up a suitably sized chunk of somethinium and place on top of the knob, repeat as necessary until the knob is WARM with due thought to not getting the somethinium SO hot as to mar any finish on the knob.

Alternatively, perhaps dry ice from your friendly welding gas distributer might work in the opposite manner?

.

[EddyCurr]

I know im stating the obvious here but it's a centrifuge so verify if the threads are left or right hand - hopefully they have enough hanging out to see or have a mark or warning label.

Well noted. Cannot say I would have had that thought first thing at this time of the morning.

Where there is one centrifuge, it seems that often another one is not too far away. If so, checking a similar unit would provide the definitive answer about thread direction.

.

[KMoffett]

Does the red arrow indicate the direction of the trunnion's rotation? If it does, then a left hand thread would work. Most lab equipment manufactures have 24/7 tech support. Give them a call.

What make/model?

Ken

[enl]

Depending on the details of the machine, my answer could go several places.

I like drilling holes for a face (pin) spanner, if it can be done without making a mess inside the unit, if there isn't a concern about the holes holding contaminants, and it can be done such that the balance won't be negatively affected. Might need to rebalance the nut or replace it with this option.

If the knurling is straight, and time permit, I would probably make an internally splined driver to grab the knurl. Mount it to a ratchet so it can only be used in the loosening direction and make it part of the machines kit. Could maybe form one faster using epoxy, instead of machining one, at the risk of getting epoxy into the machine.

If the knurling is straight and large enough pitch, an external driver could be made with three or more pins through a plate to grab the knurls. They would need to be on an appropriate circle and spaced to grab the ridges. Not a top precision job, as they can be tapered and slightly inset so they grip as the tool is pushed down over the nut. If needed, a ring around the pins will limit spread and maintain grip.

Last option would be split it (if external) or core it (if internal) and get a replacement.

[macona]

Order another nut and once it gets there do whatever you need to get the old one off.

[winchman]

I was able to remove the entire rotor assembly from the centrifuge by removing the left-handed nut on the end of the shaft, but I still couldn't get the knob off with soft jaw pliers. So I chucked the knob in the 13" Tray-Top lathe, stabilized the rotor with the tailstock ram, and used a strap wrench to turn the rotor. There are no markings to indicate that the knob or shaft nut have left-handed threads, but both do.

Fortunately the knob didn't slip in the chuck, so you can hardly tell it was gripped that way. I have no idea how anyone could have gotten it so tight. The threads are clean and undamaged.

[EddyCurr]

Well done. Imagine that produced a feeling of satisfaction.

.

[Mike Amick]

I have no idea how anyone could have gotten it so tight. The threads are clean and undamaged.

Probably trying reaalllllly hard to loosen it in the wrong direction.