Constructing the Josslin Hudson

Where users can chronicle their builds. Start one thread and continue to add on to it.

Moderator: Harold_V

Carrdo
Posts: 1444
Joined: Thu Mar 06, 2003 2:20 pm
Location: Toronto, Canada

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Carrdo »

The next operation after initially fitting all of the axleboxes is to finish the axlebox frame slots to their final shape to enable the wheels, axle and axleboxes to "rock and roll" in the frame openings so as to enable the locomotive to "track" on uneven rail.

As I have never seen any of the "trade magazines" publish any articles on this in any in depth, I am going to go into a long song and dance about this here so you will have to suffer as there are a lot of widely held misconceptions about how axleboxes should be fitted to have a free running and still "trackable" model locomotive.

The attached diagrams are taken from my construction notes developed over the years all of which are in my own handwriting so I hope they come through clearly enough.

One of the biggest misconceptions is that if you build model axleboxes to toolroom tolerances, you will end up with a locomotive that will bind and will not be able to run through curves, uneven rail, frogs, switches, etc. This is simply not true.

At the outset, I will say that one does not have to build to these standards and if one does not everything can/will be perfectly fine if recognized methods are followed (see the last sketch) but if one chooses to do so, a toolroom model can be made to run and track just as easily as with any other standard. It is all about knowing how to do it and taking the time and care to do it properly.

So to quote from my notes (which are based on the Josslin lead truck design):

Axlebox journals have to be relieved on their flanges to allow the axles to rock when running on uneven height rail. Most people file a radius here by a file and try method until a set amount of rock is achieved but I actually generate a true radius using a boring head and a horizontal boring bar set in the crosshole of a boring head.

The question to be asked is how much rock to allow? For 3/4" scale in the lead truck, I allow 1/8"- 3/16" rock such that if, say, the right hand wheel is raised by this amount above level, there is no effect on the left hand wheel (which would still track perfectly on the rail).

What radius do the journal flanges have to be cut to to achieve the above? From the attached diagrams, it is observed that if one rail is raised 3/16" in relation to the other rail, the lead truck axle will rotate in a circle whose radius is R1- approximately 3 degrees. R1 is the distance between the crown of the rails which for all practical purposes is 3-3/4" for 1/4" wide rail. It should be noted also that as the axle is rotated through the 3 degrees, the axle, axleboxes and the wheels all will tilt (and are tilted in relation to the frame on both sides as the lead truck frame will remain a level rigid box) as they are joined together by the straight rigid axle.

Therefore, you can make the journal radius slightly sharper than the 3-3/4" indicated, say 3-1/2" which is the locomotive gauge.

I haven't worked out the journal flange radius for other locomotive gauges but if you make the journal flange radius equal to the locomotive gauge, you won't be far out.

To be continued.
Attachments
Untitled-8.jpg
Untitled-7.jpg
Untitled-6.jpg
Curtis_F
Posts: 848
Joined: Mon Jan 06, 2003 10:14 am
Location: Reedley, Ca. 93274
Contact:

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Curtis_F »

Cardo,

First of all, Thank you for the updates! It's a joy to read and see your progress.


As to the question of relief for the Driver Box Jaws to allow for the axle to pivot;

On the Full size locos, there is VERY little allowance for that motion. If the angle of the axle changes; your crank pin stops traveling in a circular path in two dimensions (Up-Down, Forward-Back) and starts traveling in an ellipse path in three dimensions (Up-Down, Forward-Back, Left-Right) and twists your rods and valve gear. (I hate it when my rod gets twisted...) Which is rather tough on crank pin, knuckle joint and crosshead bearings.

I have a 1919 ALCO blueprint for a 11" x 13" driver box in front of me and it shows that for the jaws of the Driver Box there is a FLAT that extends 1" above and below the center line of the axle, then tapers 1/8" in 7" (1.0 degrees) away from the frame above and below that.

I'm sure your radius will work, but I question if it doesn't allow for too much movement which might harm the valve gear.

You also note the options of a straight taper, as used on the prototype as noted above, as well as a wider frame slot. I have to take issue with saying a wider slot could be used. Your .002" clearance on the frame width is spot on; it's a scale 1/32" which is about what the full-size engines have. Lateral Play is created with thinner flanges and/or narrowed Back-to-Back dimensions on the wheels, not with sloppy driver boxes (Except on engines with Lateral Motion Devices). The wheels need to be kept in alignment or, again, the rod and valve motion bearings will take a beating, and risk damage to the valve motion rods themselves.


Just my $0.02

Cheers,

Curtis F.
25+ Years of SolidWorks experience...now I feel old.
"You can lead a horse to water, but you can't make him drink."
"Winning an argument on the internet is like winning the Special Olympics."
www.facebook.com/HillcrestShops
Carrdo
Posts: 1444
Joined: Thu Mar 06, 2003 2:20 pm
Location: Toronto, Canada

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Carrdo »

Coming now to the actual fitting allowances for the axleboxes in the frame openings.

Front to back play of the axlebox in the frame should be limited to 0.001" - 0.0015" total; i.e. the axlebox should slide freely in the frame opening but not rattle. One acceptable test is when the axlebox is mounted in the pedestal opening it will drop down in the frame under its own weight. Lateral play between the axlebox flanges and the frame should be in the order of 0.002" on each side or 0.005" in total.

The above assumes that both the frame openings and the axlebox side slots have been machined square, are totally parallel and are truly rectangular.

In the final finishing of the 9D driver axleboxes, cuts of only 0.001" - 0.0005" were needed to achieve this. Actually, the sharpened HSS end mills I normally employ would not take such small cuts reliably so had to switch to a sharpened carbide 4 flute end mill which, surprisingly, did a beautiful job on the 9D at far below its recommended speed. I currently have a problem with the highest speeds in the mill (need to fix a slight motor pulley wobble) so have not been able, to date, to take full advantage of carbide tooling when machining difficult alloys.

The photo of the lead truck end wise shows how easily the axle will tilt (there is at least 3/16" difference here in the wheel heights) with fully radiused axleboxes. Note how the frame stays level and the wheels, axleboxes and axle all tilt. These axleboxes have been fitted, front to back, with less than 0.001" total clearance and one can not put a 0.004" thick shim anywhere between the frame and any of the axleboxes.

As you can see, it just goes that way, nothing has been forced.
Attachments
Radiused Axleboxes with Tilted Axle.jpg
Carrdo
Posts: 1444
Joined: Thu Mar 06, 2003 2:20 pm
Location: Toronto, Canada

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Carrdo »

Hi Curtis,

To try and answer your concerns.

Since I haven't yet been able to run the locomotive, will have to wait for the final proof.

Can only say that when Richard and I got the 3/4" scale Big Boy, the axle boxes were very tight, were not radiused and would not tilt. At that time, the locomotive would not track especially the tender when backing up with its five rigid axles. One of the first jobs Richard did was to radius all of the axleboxes (he filed them) and increase clearances to ..?. He also did other things to the suspension which helped the tracking.

After that the locomotive will now negotiate a 20 foot radius perfectly both forwards and backwards (which is unheard of) and every wheel will ride up and over a 3/16" thick steel plate placed on one rail without the opposite wheel leaving the track.

If any model locomotive should bind on the rods or affect the valve gear after this treatment, this one should but we haven't seen it happen. Perhaps there is allowance built in elsewhere as I haven't gone through the mechanism on the Big Boy with a fine tooth comb.

But I can ask Richard to elaborate further on what he has done.

As to prototype versus model, scale effect and track standards are totally working against us here (3.5" wheel spacing vs 4'-8-1/2" full size). With a full size wheel spacing of 4'-8-1/2" and mainline track being constructed to a much higher standard than, proportionally, what is found in model track, the full size locomotive never encounters or has to deal with the conditions under which we run our models. Or if it does, it operates at little more than walking speeds (as can be seen on the mainland China railway video which was posted here recently) One can also see on that video how much the locomotive bounces around but it still stays on the "track" likely because it is totally worn in by being worn out.

So full size locomotives do not need to have much (or any) axlebox radius due to the above. Also, I suspect that due to the axle loadings, any full size steam locomotive axlebox after 500 miles of running would be worn in and as a result would tilt to some degree knowing how hard these locomotives were worked and the minimal maintenance they received.

PS: As I haven't cut the radius' on the main axleboxes yet, if you think there is gong to be a problem with rod binding or it will affect the valve gear as Curtis fears, I am open to suggestions or a discussion on this.
RET
Posts: 960
Joined: Wed Jun 07, 2006 8:36 am
Location: Toronto, Canada

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by RET »

Hi,

Don does nice work. Referring to his comments on Big Boy, when we first got the locomotive, everything was so tight, the tender would derail on straight track. I took the tender running gear apart & filed the axleboxes so they could slide and tip freely. I also changed the coil springs (small die springs) which loked right but were far too stiff and replaced them with much lighter springs that allowed the axles to float. I also bored out the center block of the tender truck so I could fit a spring that allows the truck to float up & down. Because the bottom of the hole is unchanged, the truck still guides the tender properly and nothing shows from outside.

With these changes, I can now "throw" the tender down my outside track as hard as I can and it "rocks & rolls" as it sails along but it never comes off. As long as the wheels & axles are able to follow the track no matter what it does, the engine & cars will stay on. I did the same changes to both engines on Big Boy so they too follow the track. I didn't have to make any changes to locomotive rods or valve gear as a result of this, it just works. I made a lot of other changes, but nothing that would change the way it looks.

Richard Trounce.
Carrdo
Posts: 1444
Joined: Thu Mar 06, 2003 2:20 pm
Location: Toronto, Canada

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Carrdo »

All of the axleboxes are now initially fitted. It is a slow process at least the way I do it.

Am still thinking about how much radius to put on the main axlebox flanges.
Attachments
329 Axleboxes All Fitted.jpg
Carrdo
Posts: 1444
Joined: Thu Mar 06, 2003 2:20 pm
Location: Toronto, Canada

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Carrdo »

OK, I've decided. I am going to hedge my bets a bit. On the main axle axleboxes I am going to make the axlebox flange radius twice the gauge width or a 7" radius and also have a short parallel section in the middle. This will allow for a bit of rock but will limit the tilt of the driver axles and the crankpins to less than half of what is on the lead truck. I can see Curtis' point about wanting to keep the main rods and the eccentric crank as straight as possible and not affect the valve gear.

On the lead truck, trailing truck and tender axleboxes I am going to leave the radius equal to the gauge width so they can "rock and roll" much more as they are all independent of the valve gear.

Don't ask where the logic is - it is the best engineering compromise I can come up with for all of the competing and incompatible variables involved.

Richard told me that he can rock his B&A tank locomotive 3-4 degrees off vertical and all of the wheels on his model stay firmly on the track. He also said the the spring equalizer system on American locomotives is designed to keep the wheels on the track at all times - this is their most important function. It was his view that any bumps or uneveness in the track (excursions) are going to be temporary and the valve gear, rods and links are robust enough to be able to absorb these temporary shocks.

Time will tell in this case.

The three photos show how I set up and radius the axlebox flanges. It is important if you decide to power the boring head that it turn very slowly (about 30 rpm in my case) due to the very large sweep of the cutter. This requires good variable speed control and if you don't have it then turn the bar by hand - it works just as well this way only it is slower. Make certain the boring bar and cutter are clear of all obstructions and the cutter can reach to the bottom of the side slot without fouling anything. Ensure everything is firmly tightened to keep as much rigidity as possible.

The pointed cone in the central hole of the boring head is set to 7" on the graduated rule when the boring head slide is concentric with the body of the boring head.

Set the axlebox in the milling machine vise as shown. Again, having the axlebox truly square pays dividends here. Adjust the cross slide of the mill until the cutter just grazes equally, the top and bottom edge of the axlebox flange and check with an eye loupe. Also, as the cuts progress, one can see and measure the length of the radius at each end of the axlebox flange to see if they are of equal length.

I only took 0.005" infeed on each pass (top of flange to bottom of slot) with the table being moved up incrementally only 0.001"-0.002" per revolution of the boring bar. This is basically a slow shaving process. To avoid the cutter scoring the bottom of the slot, I turned off the power when the cutter was near the floor of the slot and swept the cutter by hand for the last shaving passes. Also, at the end of the cutting passes, the thin piece of cigarette paper (which is shown blurred in the third photo) was used to set the cutter to just not quite graze the bottom of the slot.

Only a mere 23 more flanges to go. Again, all of this is just slow careful dog work to try your patience.
Attachments
334 The First Flange Radiused.jpg
332 Setting the Boring Bar Bit Central on the Axlebox Flange.jpg
331 Setting the 7 Inch Radius on the Horizontal Boring Bar.jpg
Carrdo
Posts: 1444
Joined: Thu Mar 06, 2003 2:20 pm
Location: Toronto, Canada

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Carrdo »

Four axleboxes have now had all of their flanges radiused (16 flanges done so far) and are final fitted (from a tap to a free sliding fit).

Now the axleboxes drop down easily under their own weight and rock side to side but have no shake back to front. Can just see the gap at the top and bottom of the axlebox flanges in the photo but the flange radius only has a 0.002" or smaller gap in the center between it and the frame. I also left a small parallel section of about 1/4" long on each side slot in the center. Nothing like trying to have it all ways.

Back to front the gap between the axlebox frame slots and the frame is now 0.001" to have the totally free fit. This was done with a relatively coarse lapping compound (320 grit) and was used only because I happened to have it. The lapping process in my case was very quick and very dirty. No super fine finishing/polishing here. Just be certain to clean everything thoroughly when you are finished as any residual lapping material no matter how tiny is going to continue lapping if it is not entirely removed.

My hands and fingers now have all of the nicks cuts and bruises from encounters with all those sharp corners and edges while working the axleboxes up and down in the frame openings.
Attachments
335 Four Axleboxes with Flange Radius and Final Fitted.jpg
Carrdo
Posts: 1444
Joined: Thu Mar 06, 2003 2:20 pm
Location: Toronto, Canada

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Carrdo »

After radiusing all of the axlebox flanges and final fitting all 6 axleboxes, I thought I was nearly home.

Not so, that *!*!* 9D had more nasty surprises to come.

I wanted next to finish the axleboxes to their final, height, width and thickness in preparation for the final fine boring to fit the 3/4" diameter axles.

Even though the material had been roughed out many times and then re-setup in the surface grinder to bring everything back square, I found when reducing the axlebox thickness, the axlebox saddle arms started to spread outwards like a trumpet and very badly too. To make matters worse, where the spreading was greatest, at the ends of the saddle arms, were the thinnest sections which were previously machined out to take the leaf spring base blocks.

Obviously, there was still a great deal of residual stress left in the metal.

Hindsight is 100% and in retrospect, I now know that cast 9D must be stress relieved first before attempting any machine work. Unfortunately, few of us have heat treating furnaces and one is going to spend $$$ to have it done commercially. Fortunately, Richard has a large commercial electric furnace and so now all of the raw extra castings which I have are going there.

But that didn't solve the problem at hand. To start with, I couldn't risk heat treating the machined axleboxes for fear of them warping beyond what little extra metal remained.

I first thought about using countersunk commercial flat head cap screws to pull the saddle arms back against the leaf spring base block. Upon reflection, I abandoned the idea as countersinking would leave so little metal between the countersunk holes, there was a great risk of the metal fracturing when the screws were tightened.

Not a lot of options left but possibly, could the leaf spring base blocks be "super glued" in position and the saddle arms clamped onto them?

Using a well known commercial anaerobic glue, rated as producing an extra strong joint both in shear and in tension, the parts were thoroughly cleaned, then solvent cleaned, primed and the glue applied as per the instructions.Then the saddle arms were clamped in the vise as shown in the last photo. After 24 hours the vise pressure was carefully released. The joint lasted about 5 secs. and then popped apart. Hmmm...

After 48 hours clamping, 72 hours clamping the same thing happened, the best I could get was a joint which lasted about 5 minutes.

In desperation and frustration, more than anything, I decided to then try a cylindrical part bonder using hardened steel pins which were slightly more than 0.002" (as I just happened to find some laying about) undersized to the hole size which was only 3/32" dia. and ground slightly shorter than the clamped width of the saddle arms. I didn't think this would work due to the incredibly small surface bonding area involved but to my amazement it did. I did roughen the pin surface a bit with emery paper before applying the glue which was so watery, I thought, this will never harden.

Pure luck? I'll take it as I was now out of options.

The axlebox was then brought down to finish thickness as shown in the last photo. I fully expected the assembly would let go in the middle of grinding as the final metal thickness was only 5/64" in the area where the steel pins are but it didn't.

The pins will be removed by gentle heating until the anaerobic glue lets go. At this point, I don't care if the saddle arms spread out again as they will be brought back into alignment when the leaf spring base block bolts are installed.

Just finished the first axlebox so have to be lucky 5 more times.

The moral of the story is don't you do it this way. Stress relieve the castings first or change axlebox material to what Jack Bodenmann used on his 0-6-0 which I think was a Meehanite type of cast iron to avoid the problem altogether.
Attachments
338 Setup to Finish Grinding an Axlebox to Final Thickness.jpg
337 The Successful Method Used to Prevent the Axlebox Saddle Arms from Spreading.jpg
336 Unsuccesful Method to Prevent Saddle Arms from Spreading when Final Grinding.jpg
Last edited by Carrdo on Wed Dec 28, 2011 11:07 am, edited 1 time in total.
User avatar
Harold_V
Posts: 20227
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Harold_V »

The lesson that should be learned from this experience is the value of roughing work before taking any finish cuts.

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.
Carrdo
Posts: 1444
Joined: Thu Mar 06, 2003 2:20 pm
Location: Toronto, Canada

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Carrdo »

Unfortunately, this cast 9D material moves around (at least the castings which I have) more than one could ever imagine so it was hard to know beforehand what amount of material to leave (for roughing) and where to leave it (for roughing) to be able to bring it all back into square again for finishing. I now know the castings which I have have to be fully stress relieved first before one even makes a cut but that is experience/hindsight.

I really wanted to make the part exactly the way it is shown on the Josslin prints and I will but the price to pay for this has been steep.

Will take a few photos of the extra castings when they go into the furnace for stress relieving which is 600 degrees fahrenheit for one hour at that temperature per inch thickness of material.

Am going to have a post mortem on all of this when the axleboxes are finally finished.

Still 3 more machining operations to go.
User avatar
Harold_V
Posts: 20227
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Constructing the Josslin Hudson Part 54 (Update)

Post by Harold_V »

Carrdo wrote:Unfortunately, this cast 9D material moves around (at least the castings which I have) more than one could ever imagine so it was hard to know beforehand what amount of material to leave (for roughing) and where to leave it (for roughing) to be able to bring it all back into square again for finishing.
There are materials that display those characteristics, so you have to think accordingly. In such a case, one of the things that works is to rough, semi-finish, then finish. In such an instance, when roughing, a greater amount of material would have been allowed for removal in the following operations. You may come to discover that as a lesser amount is removed from each surface initially, the part undergoes less movement in general, so the end result is very acceptable.

The theory behind this method is that stock removal that is uniform tends to result in less movement, in spite of stresses. That theory is proven in practice when you surface thin materials on a surface grinder. If a large amount is removed from one side without removing any from the opposite face, movement is extreme---yet when the part is flipped continually with smaller cuts of similar depth taken, overall movement is reduced considerably.

I'm not suggesting that stress relieving isn't a consideration---just that there may be times when that's not an option. You can address the problem by following the guidelines I specified.
Roughing of machined parts is largely overlooked these days. The importance of roughing can't be overstressed, as your results verify.

Harold
Wise people talk because they have something to say. Fools talk because they have to say something.
Post Reply