Alan Gibson wheel wobble

[PenrithBeacon]

One of the banes of my modelling is wobbling Alan Gibson wheels. Last Saturday I was at the Woking show and I asked the blokes on the EMGS stand if they knew of a solution. I showed them a video taken on my phone of my current project and their advice was to ask Colin Seymour and to use the GW wheel press. Further conversation around the topic elicited the advice that using masking tape to secure wheels in place on the wheel press was a good idea. 

I haven't asked Colin directly but I have looked at his website, http://www.alangibsonworkshop.com/Catalogue.pdf , and there on page 36 is some advice.

Combining the two, I have this afternoon assembled a wheelset and it doesn't have any wobble!

Firstly crankpins were screwed into the wheels. Then the axle was shortened ( a little too much actually, but it will do for now). Then the axles was secured into my drill gun and a taper was filed on both ends.

 

 

Then the wheels were secured to the wheel press with masking tape

 

 

Then the wheel press was assembled and secured in the vice. The axle was offered up using snipe nosed pliers

 

 

The wheel press was the gently squeezed using finger pressure only and the B2B checked with an AGW gauge.

 

 

The wheelset was removed from the press and run along the bench, it appears to be wobble free and I'm over the moon! the down side is that I'm go have to partially dismantle the LRM Stanier 0-4-4T I'm currently building to get it up to the same standard!

 

Hope someone finds this useful

 

Regards

 

 

[Paul Cram]

How are you going to fit those to the loco?

[PenrithBeacon]

How are you going to fit those to the loco?

Actually they're the wrong wheels for a Stanier 0-4-4T. They would do for as a drop-in wheelset for a Bachmann Jinty. This afternoon was just a feasibility exercise.

 

Regards

[bertiedog]

Can I say that the tapered ends of the axle should really be a bit more delicate than the ones in the photos.

It does not need a lathe, but also should not need filing the ends, but be done using abrasive paper.

 

First give the whole stock axle a polish with a very fine grit, +1500, to give a mirror finish overall. This will not reduce the diameter, but remove any imperfection.

 

Then polish the ends with a 200 grit to a very small taper that just enters the wheel hole, use the paper stuck to lolly sticks or similar. Again finish the taper to the very highest polish.

 

The point is to leave no way for the steel to machine the hole in the plastic, it must go through without any cutting, which leaves them permanently marked in the bore.

 

Should the wheel go on too easily, then remove and roll the end under a very fine file to roughen a small area where the wheel sits, but leaving the taper. Rolling under a file edge "raises" the surface, and it will now fit.

 

If the set up is final during assembly, then add a drop of Loctite to ensure it retains. Wipe or wash away any excess Loctite.

 

Stephen

Edited September 12, 2017 by bertiedog

[bertiedog]

Also may I add if you are new to any press on wheels, not to try to use the old Hamblings wheel press for these modern plastic wheels.

The Hamblings were designed to use a pilot nosed axle that was a Swiss made precision turning that acted as a guide to press the steel into a brass bush. It does quarter quite well, but does not operate as well as the GW type, unless seriously modified.

 

Whilst the tape is a good idea for starting the push, it will then be better removed so that you can rotate the wheels as you push, evening out any off centre forces. The crankpins would have to be removed as well.

 

Stephen

[Pete the Elaner]

I have a b2b gauge like that but do not like it because I found that it allows the wheels to be gauged without being parallel to each other, causing the wheel wobble you are describing.

What gauge do you model? One of the societies may have something better.

 

Edit: I just noticed you mentioned EMGS, so I assume you model in EM?

Something like this may be better because it surrounds the axle more.

Edited September 12, 2017 by Pete the Elaner

[timbowilts]

Replying to Pete the Elaner, I have always sued the "L" shaped b2b gauge in my EM modelling. The trick is not to use it as in the photograph but to use it so that the axle is inside the "L" shape.

 

If you do that then the wheels must be parallel because you are measuring the b2b in two planes at once

 

Tim t

[micklner]

Also may I add if you are new to any press on wheels, not to try to use the old Hamblings wheel press for these modern plastic wheels.

The Hamblings were designed to use a pilot nosed axle that was a Swiss made precision turning that acted as a guide to press the steel into a brass bush. It does quarter quite well, but does not operate as well as the GW type, unless seriously modified.

 

Whilst the tape is a good idea for starting the push, it will then be better removed so that you can rotate the wheels as you push, evening out any off centre forces. The crankpins would have to be removed as well.

 

Stephen

 

How would you re fit the crankpins afterwards ? The GW uses the pins for alignment/quartering as well ?

[bertiedog]

They should unscrew after having them in place to push on with quartering. But anyway the quartering should be checked and adjusted after the wheels are in place, they will move under moderate force with fine Jewellers round nosed pliers.

 

It all depends on the chassis they are to go in, drop in, with bearing blocks, as per the real thing, or horizontal split, with or without bearings, as per most commercial models, or pressed on into a chassis that has complete bearings, more common on older designs.

 

One solution does not fit fit all.... the order of operations has to be adjusted to suit.......and may involve establishing the quartering in the press to start with, then removing the pins to continue an accurate press......then replacing the pins afterwards.

 

However with a pin system that requires access to the back of a wheel to secure, then they have to remain in place if fitted to a solid frame.

 

There are several makes of crankpin that will fit Gibson, my favourite would be the Markit De-Luxe type, which can fit most wheels. The Gibson design is OK, but check the spacers are concentric, there was a batch of bad ones a while ago, only slightly off centre, but enough to make fitting the rods foul the holes. This was mentioned to Gibson, and they replaced all faulty owns.

 

By the way, if you have a press drill, a column drill press, this is perfect to press Gibson or Ultrascale on to the axles. The axle is placed in the chuck, and the axle pressed gently home in to wheel, (may need packing to allow for bulging boss).

The second wheel is started on the axle taper, and the lot pressed home with a blank axle, made thinner than 1/8 by emery paper, acting as a pin punch held in the chuck. With polished axles and fine tapers the lot slide home with complete accuracy. A small piece of perspex is used to check quartering afterwards.

 

Stephen

[bertiedog]

I have a b2b gauge like that but do not like it because I found that it allows the wheels to be gauged without being parallel to each other, causing the wheel wobble you are describing.

What gauge do you model? One of the societies may have something better.

 

Edit: I just noticed you mentioned EMGS, so I assume you model in EM?

Something like this may be better because it surrounds the axle more.

The Gauge should not matter or cause any wobble, it is the pressure on the wheels from the jig, or jaws of the vice, or hand pressure, that is behind the wobble. The gauge should only be used for checks, not in place during the press. Some gauges are made to allow them to check the quartering as well, these allow the axle to sit in a groove at the right angle. In an engineering shop it is just not done to use a gauge to check things by pressing on till the gauge is gripped, it is repeatedly fitted and un-fitted till you feel resistance as it is inserted. This is usually the correct BB.

 

Any more pressure may close the grip, but may also bend the plastic spokes, which when the gauge is removed spring back and give a slightly false BB.

 

Also a very basic point, always check any make of wheel has no trace whatsoever of flash or sprue marks on the rear surface, it may look clean, but check with a steel rule.

[Gordon A]

I have added plastic card packing to my GW models jig so that the wheel rims are also supported.

I have found that this works quite well.

 

Gordon A

[PenrithBeacon]

I have a b2b gauge like that but do not like it because I found that it allows the wheels to be gauged without being parallel to each other, causing the wheel wobble you are describing.

What gauge do you model? One of the societies may have something better.

 

Edit: I just noticed you mentioned EMGS, so I assume you model in EM?

Something like this may be better because it surrounds the axle more.

It's best if you have two of the 'L' shaped b2b gauges. I have but one was omitted from the photo. They are AGW but are also supplied by the EMGS and S4 Society .

Regards

[PenrithBeacon]

The Gauge should not matter or cause any wobble, it is the pressure on the wheels from the jig, or jaws of the vice, or hand pressure, that is behind the wobble. The gauge should only be used for checks, not in place during the press. Some gauges are made to allow them to check the quartering as well, these allow the axle to sit in a groove at the right angle. In an engineering shop it is just not done to use a gauge to check things by pressing on till the gauge is gripped, it is repeatedly fitted and un-fitted till you feel resistance as it is inserted. This is usually the correct BB.

 

Any more pressure may close the grip, but may also bend the plastic spokes, which when the gauge is removed spring back and give a slightly false BB.

 

Also a very basic point, always check any make of wheel has no trace whatsoever of flash or sprue marks on the rear surface, it may look clean, but check with a steel rule.

Not quite sure I'm following this.

Agree with cleaning up a wheel before proceeding. But having said that I'm not sure I can agree with the rest, sorry.

Surely the whole point of using the press is to get the wheels square and quartered first time, every time. It makes no sense to me to adjust the wheelset after it comes out of the tool.

Perhaps I'm misunderstanding.

Regards

[PenrithBeacon]

Can I say that the tapered ends of the axle should really be a bit more delicate than the ones in the photos.

It does not need a lathe, but also should not need filing the ends, but be done using abrasive paper.

 

First give the whole stock axle a polish with a very fine grit, +1500, to give a mirror finish overall. This will not reduce the diameter, but remove any imperfection.

 

Then polish the ends with a 200 grit to a very small taper that just enters the wheel hole, use the paper stuck to lolly sticks or similar. Again finish the taper to the very highest polish.

 

The point is to leave no way for the steel to machine the hole in the plastic, it must go through without any cutting, which leaves them permanently marked in the bore.

 

Should the wheel go on too easily, then remove and roll the end under a very fine file to roughen a small area where the wheel sits, but leaving the taper. Rolling under a file edge "raises" the surface, and it will now fit.

 

If the set up is final during assembly, then add a drop of Loctite to ensure it retains. Wipe or wash away any excess Loctite.

 

Stephen

The courseness of the angle was deliberate. The object was to locate the axle in the centre of the hole but also to make absolutely sure that non of the plastic was gouged out of the hole. As you rightly say such gouging is a major cause of wobbly wheels. Been there done that.

Your comments re surface finish are interesting. I think that an excessively smooth finish can contribute to the wheel shifting on the axle. It's all a matter of degree. A wheel does need to be keyed onto the axle

Regards

[johnarcher]

If the set up is final during assembly, then add a drop of Loctite to ensure it retains. Wipe or wash away any excess Loctite.

 

Stephen

Do you mean add Loctite to the face of the wheel after it is fully home, when you are sure all is quartered OK? Does it penetrate?

I did once try combining Loctite with pushing wheels home with a GW jig, and Loctited the steel retracting pin in the jig into its hole.

[bertiedog]

It is near impossible to get a good finish to ruin the push on action of both the Gibson or the Ultrascale types, whose mounting standards are about the same. Such fine grit polishing does not affect the diameter beyond a few microns here or there, maybe about 1/10 thou at most. The rest of the grip is due to the flexibility of the plastic, pushed home forcing the boss to expand and grip.

 

What is not needed is any kind of digging it to the plastic at all, as once stared it is a permanent scar in the bore.

I always add a tiny spot of Loctite anyway, and this air exclusion glue will do the final fixing touch job.

 

However in the real world there are makers tolerances, the axle may be a bit undersize or over, and the centre moulded bore may be a little out of diameter, and also contains"draw", a minute tapering of the hole to allow the die to withdraw after moulding.

 

So if the axle is a tight fit as pressed, then all is well, if it goes to loosely then the axle end requires raising the surface with a file, the axle is placed on a cutting mat, and a fine Swiss file is pushed down on the end area, and rolled back and forth till a nice matt micro knurled look appears. It will be about a thou larger than before and still slide on, but also has a good grip with the loctite.

 

This assumes the axle and the bearings drop into a kit chassis or home produce item, where such a chassis accepts drop in wheelsets. Other chassis may need a different approach altogether, to get the order of assembly right and logical.

 

On a set, to perfect the quartering, any axle with a push on wheel will shift under moderate force, but still grip correctly for normal use.

 

If all the axles have had the micro knurling done on them, then such correction is near impossible, so be more careful and use the quartering jigs with care during pressing on!

 

I must on reflection done many thousand of  wheelsets, all makes and types, such as push on Romfords, and screw fitted Hamblings, plus screw retained Gibson and Ultrascale, plus hundred of corrected K's wheels.

 

Most are basically true to use, but it is very easy to loose course and end up with wobbles or eccentricity large enough to affect the smooth running of the locomotive.

 

No such thing as an off the shelf loco wheel thats perfect, but in my opinion the Ultrascale or Gibson are the best, if the boss in the middle is bored out in a lathe, and a brass bush is fitted with loctite and then the bore is done by boring out and reaming the hole.

The hole is smaller than the 1/8th axle and the axle end is reduced to match. The wheel slides on easily, and the lot retained by a 10ba bolt into to the axle end. They are nipped up to get grip with loctite, which remains liquid long enough to set the quartering.

 

But this requires a lathe, and most modellers do not have one, so using the tradition push on method remains the answer for most people.

 

Stephen.

[bertiedog]

Do you mean add Loctite to the face of the wheel after it is fully home, when you are sure all is quartered OK? Does it penetrate?

I did once try combining Loctite with pushing wheels home with a GW jig, and Loctited the steel retracting pin in the jig into its hole.

Only the tiniest drop inside the bore, placed there by a pin or wire, just a trace is all thats needed, as it sets when air is excluded after pressing home, any left must be washed away to ensure it does not get into places it should not be! No need to hurry with Loctite, it remains fully liquid until the air is not in contact. In theory superglue will work, but you may get it set too soon, and the grip on the plastic is not quite as good.

Also Epoxy glue will work, but it should be a long set high strength type, not the softer five minute type.

 

Stephen

[bertiedog]

By the way, the real danger to smooth running is not side to side wobbles, but eccentricity, which ruins the electrical pick up reliability. Even a few thou can force an 060 to rest on the track on just two wheels, rocking slightly as it stands still.

 

One old K's I repaired in the trade had almost 1/16 inch wobbles on all axles, looking horrendous when run up side down, but just glide along the track in practical use!

 

But one wheel out of concentric, and all smooth running is impossible as it thumps it's way along the track. It also ruins the pulling power.

Stephen.

[Jol Wilkinson]

See Tom Mallard's article on using Sharman Wheels on LSWR Black Motor 0-6-0s in MRJ 165 for some practical advice. While a lathe is useful you can gently ream the wheel bore if you have a pedestal drill (rotate by hand).

Ultrascale Wheels have the best reputation for concentricity/run out of those currently available, partly achieved by a less tight fit on the axle. However that often means that the wheel needs pinning to the axle, something to be taken into account if you bore out AGW or SW wheels.

[johnarcher]

Only the tiniest drop inside the bore, placed there by a pin or wire, just a trace is all thats needed, as it sets when air is excluded after pressing home, any left must be washed away to ensure it does not get into places it should not be! No need to hurry with Loctite, it remains fully liquid until the air is not in contact. In theory superglue will work, but you may get it set too soon, and the grip on the plastic is not quite as good.

Also Epoxy glue will work, but it should be a long set high strength type, not the softer five minute type.

 

Stephen

Thanks for the reply. However I thought that's what I did - a drop in the wheel bore, which then got onto the locating pin when placed on the jig, and, as the wheel was pushed on, the pin retracted with a little Loctite on it.and stuck. Perhaps I misunderstand what you say. I'd like to get this clear, as I have some to do soon (I hope) and would rather not try to pin them.

[Guy Rixon]

Only the tiniest drop inside the bore, placed there by a pin or wire, just a trace is all thats needed, as it sets when air is excluded after pressing home, any left must be washed away to ensure it does not get into places it should not be! No need to hurry with Loctite, it remains fully liquid until the air is not in contact. In theory superglue will work, but you may get it set too soon, and the grip on the plastic is not quite as good.

Also Epoxy glue will work, but it should be a long set high strength type, not the softer five minute type.

 

Stephen

 

 

If the wheels and axles were prepared to be a sliding fit and secured only by the Loctite, do you think that it would give a good result. I can see that it wouldn't work with a GW press as the retainer would get everywhere and ruin the press. But generally, would the wheel stay in position?

[Jol Wilkinson]

John,

 

a friend has pinned wheels by cutting a slit in the end of the axle at about 30 - 45 degrees (across half the face of the end) and drilling into the wheel centre along that. A piece of n/s or brass wire is then used to pin the wheel to the axle. 

 

If you need the remove the wheel the pin will slide out of the slit. 

 

I am about to try that, on my next loco.

 

Jol

[bertiedog]

If there is a full set to do, and drop in bearings, and Ultrascale, then polish the axles, and add the micro knurled ends, dab of loctite and push home in quartered position, job done.

Same for Gibson, but they are a tighter fit than Ultrascale anyway.

With any pinning leave till running tests, then drill and pin.

 

Stephen.

[Torper]

I use Loctite retainer with a GW wheel press (together with plastic card packing and an L shaped back to back gauge for that matter).  Yes, the retainer will gum up the press if left, so immediately after taking the wheels out of the press I clean it up and all is then well.

 

DT

Edited September 13, 2017 by Torper

[NFWEM57]

On 12/09/2017 at 20:37, bertiedog said:

It is near impossible to get a good finish to ruin the push on action of both the Gibson or the Ultrascale types, whose mounting standards are about the same. Such fine grit polishing does not affect the diameter beyond a few microns here or there, maybe about 1/10 thou at most. The rest of the grip is due to the flexibility of the plastic, pushed home forcing the boss to expand and grip.

 

What is not needed is any kind of digging it to the plastic at all, as once stared it is a permanent scar in the bore.

I always add a tiny spot of Loctite anyway, and this air exclusion glue will do the final fixing touch job.

 

However in the real world there are makers tolerances, the axle may be a bit undersize or over, and the centre moulded bore may be a little out of diameter, and also contains"draw", a minute tapering of the hole to allow the die to withdraw after moulding.

 

So if the axle is a tight fit as pressed, then all is well, if it goes to loosely then the axle end requires raising the surface with a file, the axle is placed on a cutting mat, and a fine Swiss file is pushed down on the end area, and rolled back and forth till a nice matt micro knurled look appears. It will be about a thou larger than before and still slide on, but also has a good grip with the loctite.

 

This assumes the axle and the bearings drop into a kit chassis or home produce item, where such a chassis accepts drop in wheelsets. Other chassis may need a different approach altogether, to get the order of assembly right and logical.

 

On a set, to perfect the quartering, any axle with a push on wheel will shift under moderate force, but still grip correctly for normal use.

 

If all the axles have had the micro knurling done on them, then such correction is near impossible, so be more careful and use the quartering jigs with care during pressing on!

 

I must on reflection done many thousand of  wheelsets, all makes and types, such as push on Romfords, and screw fitted Hamblings, plus screw retained Gibson and Ultrascale, plus hundred of corrected K's wheels.

 

Most are basically true to use, but it is very easy to loose course and end up with wobbles or eccentricity large enough to affect the smooth running of the locomotive.

 

No such thing as an off the shelf loco wheel thats perfect, but in my opinion the Ultrascale or Gibson are the best, if the boss in the middle is bored out in a lathe, and a brass bush is fitted with loctite and then the bore is done by boring out and reaming the hole.

The hole is smaller than the 1/8th axle and the axle end is reduced to match. The wheel slides on easily, and the lot retained by a 10ba bolt into to the axle end. They are nipped up to get grip with loctite, which remains liquid long enough to set the quartering.

 

But this requires a lathe, and most modellers do not have one, so using the tradition push on method remains the answer for most people.

 

Stephen.

 

Do you have any pictures of the the modification you describe in the final part of your message?  Picture paints a thousand words.