6" Burrell Gold Medal Tractor

June 2009: To date I have only machined  the Chimney Saddle but I am now making a determined effort to get on with the front wheels.   As the drawings are not yest available I had first to generate my own Front Wheel Drawings. 

Machining the slots in the front hub

Machining the spoke slots was the first real job I have preformed on my Bridgeport since I acquired it in the spring.   I cut each slot with only 3 passes of the cutter.  The first cut went down the middle followed by to finishing cuts to set the exact width of the slot.  I have fitted professional digital scales to the X, Y & Z axis and this enabled me to reset the cutter to precisely  the right place for each finishing cut with any difficulty

The job proceeded so quickly I nearly forgot to align the slots on the second side to the first.  To do this I placed a spoke in a under slot and clocked it so that it was aligned with the X axis and the zeroing the rotary table to this position.

The T rings needed welding to the rims and I knew my MIG welder was not up to the job.  I have previously had a downer for stick welding as the results I have seen look dreadful.  At the club the other day an experience member was welding a lot of poles for another member and I was surprised at the quality of weld he was archiving.  I even had a go and was surprised to be able to do a half decent weld.  Yet another club member had a large oil filed welder surplus to his requirements and I was able to acquire this at modest cost.

However, once home I found it hard to produce a decent weld and most of my welds were a mixture of slag and metal.  At the club the equipment was set up just right and I had an expert to guide me.  I took one of my test pieces to the club and asked what I was doing wrong.  I was expecting general advise but was amazed when I was told exactly what I was doing wrong.  I was allowing too big an ark to form.  I was going for the biggest ark I could get rather than holding the rod as close as possible to the job (without touching).

Welding the joint in the rim went reasonably well.  I thought I had achieved a perfect weld at the first attempt but when I chipped away the slag I was disappointed to find I had wondered off course at the beginning of the weld.  It was a straightforward job to grind out the mistake with the angle grinder and reweld.  I include the following sequence as encouragement to other non/novice welders.  Oh and the experts out there may be able to tell me what else I am doing wrong.

At first sight the weld looks good The slag looks nice and uniform	The slag even comes off with one hit To revel a problem
Problem area ground out	And rewelded
Better but the molten metal ran off the edge	Cleaned up ready for the T Ring

joint in T-ring
Is there really that much extra?

Pushing the ring into alignment
Round an round I went until no more movement

I wish I could report that welding the T-ring to the rim went smoothly but it didn't.  I had been previously trial fitted the T-rings to the rims and I expected them to fit without problem.   Since welding the rim there was no longer any give and the ring would no longer slide in.  Little Samson supply the rims and T-rings ready rolled and my gut feeling was they were about the right size despite not seeming to fit.  This is because I could see daylight in places between the ring and the rim.  I dressed the joint in the ring but did not weld it at this stage in case I would need to reduce the diameter by removing metal from the joint.  In order to stop it springing open I held to butting ends together with a tool makers clamp.  I tried to hammer the ring in place but it was a hopeless task.  I managed to get the rim in at a slant angle but could not get it square.

Expecting the ring to be a loose fit I had previously made up spacers to rest it on during welding.  These were half the width of the rim less the thickness of the rim.  I used these in a different way than expected.  I used these as pushers/stops to push the ring into alignment between the jaws of the vice.  I did this in stages going working round the rim.  It was quite handy that my spacers rolled as I went round countless times until the ring was exactly central.

The gaps between the ring and the rim had closed but I was aware that I had built up considerable stress.  I was concerned that the ring might spring out during welding.  It was necessary to make a judgment call as the forces were ensuring the ring was against the rim.  I decided to press but see below.

While I could now do reasonable butt welds some of my T-ring welds were dreadful. I had to grind out and start again in several places.  I have completed one side of one rim.  The Rim is strong an true but I shall now reflect whether the finish is good enough.  I am just off to my club and hope to talk to my welding consultants.

T-Ring weld in region of ring joint Not perfect but is it good enough?

Welded T-Ring

Riveting

Rivet bolts supplied by ModelWorks for the 4" Burrell

The next idea I cam up with was to weld the palms of the spoke to the rim and glue dummy rivets in position.  I believe this could work for the back wheels but the palms of the front wheels are too visible and the weld would be obvious.  I have not explored this option further as the hard nut to crack is the front wheels.  While I Whissendine I chatted to Dave Bennion and discovered that he had used the rivet nut principle on his back wheel.  My thinking is that I can thread a real rivet and use a 5/16" nut on the inside.  If I round the edges of the nut using a form tool prior to fitting and bearing in mind it is almost hidden by the T ring, I should get away with it.

Back to real rivets; the issue I have with forming  the head is noise.  I could not stand the noise myself and I am sure I would fall out with my neighbours who are only feet away.  Chatting this though at the club it was suggested to me that I would use and over size G clamp, with snaps to form the head by compression alone. I was told of successes forming heads on 1/4" rivets and it was considered possible for 5/6" rivets.

Press made to test principal of forming head by compression alone


Test results together with snaps


All heads have equal volume

The forces required for form a head by compression alone without hammering are very large.  Since this would require a substantial fixture I wanted to prove that the basic idea worked before committing myself by building a test press.   By restricting the access to the riveting space I was able to take the load around both sides of the rivet thereby dispensing with the need for very thick metal.  My test press was constructed from a length of M25 studding operating through a M25 nut.  An work platform was suspended some 40mm below the nut with the aid of 50 x 6 mm side plates welded in position.  The end of the M25 studding was bored to allow the inserting of a range of rivet snaps that I would use to form the rivet.

In the 1/4" riveting example I was told of the head was formed in one go with a singe snap.  I though this was too ambitious and decided to use intermediate snaps to progressively form the head as described in Edwards book.  My first attempt was a failure , the rivet bending even before I had the shaft fully home.  My rivets were a tight fit in the holes I had drilled and it needed a special snap to close up the job before starting to form the head.

My second attempt was better but my finishing snap severely distorted itself because I did not allow sufficient metal above the dome.  With an improved finishing snap my third attempt was better.   Comparing my formed head against the original preformed head it looks as if my head is too small.  Surprisingly it is the preformed head that has got larger!  I believe this is because the snap I rested the preformed head in had an incorrect profile.  I had used a 5/16" radius D bit to form the hollow in the snap but I now know I should have used a 0.276" radius.  I think somewhere along the way I had used the proportions for an aliminium rivet which are different.

In an attempt to design better snaps I have produced a drawing of a generic steel rivet.  The drawing is for a 1" rivet but all dimensions are intended to be multiplied by d the shaft diameter of the rivet.  In my case 0.313 thus the height of the dome is 0.75 x d,  0.234".

With 1.43d poking out I plan to half form the base of the head.  To simplify construction of the snap I have allowed for it to be drilled and it is thus shown with a drilled rather than flat top.  The second operation is intended to fully form the base keeping to the double cone profile.  Finally I will finish the head off with a domed snap.  I have included the flat toped cone for reference and do not intend to make a snap of that profile.

Wheel Building

Crown on wheel caused by distortion during welding

When I checked that the wheel rims were at right angles to the T-ring I had a shock - they whern't!  I feared that the T-Ring was bent and thus wound not provide a true surface for the spokes.  On checking I discovered that the T-Ring was flat and true however the rim had distorted giving the wheel a crown. When I fitted the T-rings they were a "hammer fit" inside the rims and as the centre of the rim got red hot during welding , the T-ring pushed the centre of the rim out.  With hind sight I should have made the T-rims a loose fit in the rim so that there was not considerable in built tension to be released during welding

Setting the DRO datum to hub centre using a dial gauge


The Bridgeport's head wound out to reach over the 18" diameter rim

I plan to build the wheels on the table of my Bridgeport.  Not only is this flat it will serve as a coordinate drilling machine.  First I machined some packing pieces so that I could bolt the hub at  the right height so as to line up with the desired position with the rim bolted to the table.  When bolting the hub to the table I took care to align a spoke slot with the Y axis of the mill.  I then moved  the table hub so that it was directly under the milling head  and zeroed my DRO (digital readout) to this position.  All setting operations were accomplished using a dial gauge held in the chuck.  I find a dial gauge indispensable and have just brought a second so that I have one in each workshop.  With a long arm on the dial gauge I was able to position the rim so that t it was concentric with the hub.  Since my rims are not perfectly round I spend some time finding the best compromise position before bolting the rim in place.

The first operation preformed was drilling the holes for the rivets in the rim.  I generated a special purpose drawing from my CAD model that gave me the hole XY coordinates relative to the hub centre.  With my DRO set to this datum it was a fairly simple task.  The only complication was half way through I had to wind out the Bridgeport's head and re-pickup my datum in order to extend the Y travel of the table to reach to the front of the wheel.  I should not complain as the Bridgeport + DRO makes a marvelous coordinate drill machine.

Jig used to precisely bend spokes


Drilling centre spoke fixings

Edward in his book recommends using a jig to bend the spokes.  I made something similar using some C section steal from the "scrap box".  The short end of the spoke is clamped to the bed of the jig on a spacer so that a flat piece of metal can just be slid under the spoke.  The flat piece of metal is them levered up producing a crisp bend in the desired location.  I soon discovered, by measuring the distance from the bed, the extent of bend required to produce a perfectly fitting spoke.  Once wrinkle encountered was that when the second bend is preformed the first bend gets in the way.  This is why my flat plate is cut short with an extension leaver on the side.

The palms of the spokes will eventually be held in place with 5/16 rivets but as a temporary measure I used M8 bolts.   In theory 8mm  is 2 1/2 thou over size but I found that my stock M8 bolts provided a nice snug fit.

After the first 5 spokes were fitted I turned my attention to the hub fixings.  Each spoke is held in place by a M6 CSK screw plus a 6mm locating dowel.  Once again I coordinate drilled these holes using dimension extracted from my CAD model.  My DRO can store up to 99 relative zeros and I set a zero at each hole position.  This was useful as each hole required several tool changes to complete (eq centre drill, tapping drill, clearance drill).  I was then able to complete all operations with a drill before changing to the next.  Whilst drilling the first hole I set the depth stop which meant that subsequent holes could be quickly drilled using power down feed.

When the wheel was turned over for the second side I machined packing pieces for the rim of just the right thickness.  Once the DRO datum had been set drilling the hub was a breeze as I had prevously stored the hole positions.

This is the first time I have done a major job an my Bridgeport and I love it!  I found that the spokes fitted surprisingly easily (every thing lined up) and I even believe that my spokes are interchangeable though I have no intention of putting this to the test.  All  I need now is a larger Bridgeport so that I can make the back wheels the same way.

Front wheel after fitting spokes

10 Tonne hydraulic press used to apply force to riveting dollies .. but is it enough?


Initial test with press and new dollies

I order a 10 Tonne Sealey hydraulic press three weeks ago and it has only just arrived.  I searched the Internet to obtain the best deal but unfortunately when things when wrong their customer service was lacking. The company I found were basically order takers who had access to generous discounts offered by Sealey.  The press would be dispatched direct from Sealey.  When no press arrived and after several lengthy phone calls, I discovered  that the press had been lost in transit by the carrier.  Fortunately Sealey agreed to send a replacement and this arrived 1 day later than expected.  While the intermediary offered me the best price going through theintermediary did not allow me to track my order online despite several request to be phoned back with a tracking number.

I bored a 1" diameter hole in the mandrel of the press to the riveting dollies I intended making.  For speed of production I made these from 2" lengths of 1" diameter free machining bar stock.   Ideally I should have used some form of exsotic steel that would better withstand the pressures.

I made four dollies as dollies as follows: -

• plain 5/16 deep hole used to compress the plates and push the rivet home
• Steep cone used to initially expand the base of the rivet.
• Flat cone used to further expand the base of the cone
• Finished shape (round top) used to form the final shape

Before starting on a wheel I thought is wise to do further test rivets.  I was disappointed with my first test and concluded that the press was not exerting enough force.   The top of the head was fine but I was unable to  form the bottom of the head where the rivet meets the underlying metal.  To try on over come this I used a flat length of bar to push down the top of the rivet.  I then found that my finishing snap was able to push down the outside of the head but try as I might I could not remove the flat.  I initially though that I had not allowed enough metal to form a full head but further experiments proved that this was not the case.

Final test rivet looks pretty good
The head I formed is at the top

I believe the correct solution to my dilemma would be to get a 20 tonne press decided to try on for the head in two goes.  I had show that pushing down on the centre of the head pushed out metal so that I could form the head outer.  Rather than use a flat dolly for this I made a new "finishing dolly" and then proceeded bore away the outer edge in the shape of a cone.  In use this dolly would apply all its force in the centre.  I then made yet another "finishing dolly" but this time drilled out the centre using a 3/8" drill.  Inorder to avoid forming too big a step the sentre drilling is only 1/32" deep.  In use this dolly applies all of its force on the outer edge of the rivet.
To overcome the limitation in force available I plan to use five as dollies as follows: -

• plain 5/16 deep hole - compresses plates and pushes rivet home
• Steep cone - initially expands  base of the rivet.
• Centre finishing - finishes centre and pushes metal out
• Outer finishing - finishes outer edge
• Finished shape (round top) used to form the final shape

By using the center, outer & finishing dollies several times I was able to form a acceptable head on the test piece shown on the right.

The first head on the actual wheel did not go according to plan

I concluded that the problem was being caused because the top dolly was not remain vertically above the bottom as presure was applied.  To overcome this I decided that I would have to make a jig to hold the dollies in allignment as presure was applied.

I have now sucessfully done 4 rivets and it is slow going as several changes of the riveting dolies are required to form each head (between inside & outside finishing).  I hope to get the time per rivet down to ten miniutes.

The jig used to hold the dollies in alignment	General view of riveting setup
The jig in use in the press	I am now happy with the formed head

.........