Home‎ > ‎Part Index‎ > ‎

Connecting Rods

A website dedicated to the construction of an accurate 1/2 scale replica of a 1937 Aero Douglas Motorcycle


The connecting rods on this scale model are made from 3/8" thick gauge plate - not because it was required but because I had some the right size and the full size ones are partially hardened cast steel.  The big end has a single row ball bearing and the little end is bronze, similar to its full size counterpart.  The full size rod big end bearing is made from many 1/4" X 1/4" individual rollers held in a aluminium cage which ran directly on the hardened bore of the rod, this I felt was too much hassle as the rollers would be 1/8" X 1/8" and having to either find some or make some and then harden the big end without it all warping or twisting would've been extremely time consuming.  I therefore used a commercially available ball race, easy.


Step 1

The full size components were measured and then a rough fag-paper sketch drawn, it was decided that a ball race would be used in the big end and a needle-roller bearing in the little end.  This was changed later to a bronze bush as the needle roller bearings weren't suitable due to their size, this means that the little end is slightly larger in diameter than scale - even though its wrong, it can't be a bad thing.  A blank was then held in the vertical milling machine and the centres for the holes were marked a 2-5/8". 

Still in the milling machine, metric drills were used in stages to open up the holes to within a few thou of required size (approx 9/16").  Once the bores were close to size, a set boring head was used to take a light skim out of each bore, at the same setting, to get them the same size.  This means that when placed on a jig, they can be reversed or turned over accurately.  Both of the blanks were done like this.


A jig was required to hold them in the milling machine for profiling the outside.  A steel plate - which had been used many times before for this kind of thing, was counter-bored to accept accurately made buttons the correct diameter of the holes in the gauge plate blanks.  This allowed the blank to be slid on very accurately and then held there with some M6 socket head allen bolts through stepped caps. This then allowed all the machining work to be done without machining into the fixings or the jig.  (The jig can be seen further down the page)



Step 2

Using the scaled drawing, a CNC programme was constructed to run around the outside and produce the profile.  The profile can be done using a rotary table and some calculations, but is time consuming and I only have so many hours to play.  This procedure is described in most of the model engineer handbooks and workshop manuals around.

The first thing i did was to tell the operating software that the cutter i was using (8mm end mill) was larger than it actually was, this automatically leaves extra material on making it oversize.  The program i used required me to manually, after each pass, lower the cutting plane so that it would take a deeper cut.  I worked my way down approximately 20thou at a time until I had created the profile.  I then told the operating software that the cutter was its correct diameter and tool a final pass with plenty of coolant at full depth - the results were good to say the least. 

I now had two profiled conrods with two identical bores.  The next stage was to get each portion to its correct thickness as the full size counterpart has quite a complex shape to it.

Unfortunately I was so engrossed in doing all of this I forgot to take any pictures, I must remember to take my camera everywhere!



Step 3

The rotary table was then brought into action.  The different cut outs and radii at different levels were drawn into a CAD package and the angles were calculated, this made this job particularly easy.  A ball nosed cutter was used to form the 'strap' that went around the big end and a 3/8" old slot drill was ground by hand to give an approx 60thou radius in the corner where the shank of the connecting rod joins the big end at a lower level. 

The jig plate allows the components to be held accurately no matter how many times you remove and replace the connecting rod blanks (as long as they haven’t twisted, although gauge plate is meant to be very stress free and not do this so much) and it also allows you to hold them vertically for thicknessing or horizontally for profiling accurately. 

The next stage was to machine the thickness of the shank, this was done on the jig plate again to ensure accuracy and because it's just so easy! 

A new end mill was used of a specific diameter to create the run-out at the little end.  Climb milling gave the best final finish although you must make sure your machine is not going to grab the job - applying the table locks lightly can help. 

The readings on the dials were taken and safety distance stops on the table set to make it easy to repeat, as the other side was done by turning the blank over and repeating, same with the twin blank.  It was then removed and lightly deburred using a fine needle file and some fine emery paper.  It's difficult not to round corners whilst deburring but it looks allot nicer if you don't.



Step 4

Once the blank had been milled to thickness it was replaced on the cnc milling machine table ready for the pocket either side of the shank, making it 'I' in section.

Once again, an slot-drill was taken to the off-hand grinder and some radii ground onto the corners without turning it into a ball-end mill.  These edges were always cleaned up with a stone under a lupe and seemed to produce just the right shape.  The flats remaining on the slot drill means that I could clean the middle out with the same cutter and not worry about any join in the cut.


Once this had been done to each side of both blanks, the jig was returned to the manual milling machine to mill in the run-out of the pocket at the little end of the rod.  This was done by placing the jig on a tilting angle plate, a ball nose end mill and allot of black marker pen to see when the cutter was blending in with the side of the pockets. 


It came out really well although it was a bit tight on space at times, those counterbored caps used to hold it down were worth the extra time making them.



Step 5


At this stage it was time to bore out the big end to fit the bearings.  The little end was already the correct size for the bronze bush but the larger end needed to be 21mm for the ball race.  Alot of the meat was removed by lightly holding it in the 3-jaw chuck but as the big end got thinner in section, this method of holding became unsuitable. 

A ring clamp was used here to complete the bore to the correct size.  The ring clamp consisted of a piece of steel, larger in diameter than the big end with a counterbore made only slightly larger than the required bore.  This face had a selection of tapped holes which matched clearance holes in a 1/4" thick ring, again with a bore just slightly larger than that required.  This meant that the connecting rod could be held square and me made to run true very easily, accurately and safely without the worry of squashing or miss-shaping the part.  I also use this type of clamp on eccentric straps for steam engines etc. with great effect.  


Once the big end bores were completed on both rods, the 21mm OD ball races were pressed in.  The bores were approx 3-4 tenths under the size of the bearing to give a nice fit without crushing the bearing.  I could not get a bearing the diameter and width required in one, therefore two spacer rings were machined to be a nice push fit either side of the bearing.  These acted as an oil trap and accurately spaced the bearings. 


The crankpins are part of the inner web of the crankshaft and are 9mm diameter and the bore of the bearings are 14mm.  Two top-hat bushes were turned up to a good fit on the crankpin and bearings whilst being 3 thou aside wider than the connecting rods.  This means that the crankshaft can be tightened fully without gripping the connecting rods and still leave a small gap for oil ingress. 


The full size counterparts are cast steel and have a rough sandy finish, so to imitate this, the rods were sand blast with silicone sand.   Firstly, as the bearings were already inserted and it was VERY important that none of the sand got into them, 8 stepped caps were turned from steel and drilled to accept M4 bolts.  These were used when sand blasting and weren't removed until they had been cleaned deeply.  The sand blasting removes all the oil and grease from the surface so after it was cleaned it was oiled to prevent rusting.  The surface finish looks very good and only took 30 seconds to achieve on each rod.  I will also be using sandblasting to achieve a cast look on the crank cases which have been machined from billet aluminium.




Stage 6


The completing part of the connecting rods are the little end bearings.  These are bronze full size and are easy to replicate as bearing bronze (not the red stuff) is a commonly used material.  Brass should not be used as a bearing material as it's wear characteristics make it only suitable for light loads or even static models.  The bronze was turned in the lathe, bore first, then the outside, plus 5-6tenths oversize to give it a suitable press fit (normally around 1 thou per inch for a good tight fit).  They are also 5 thou wider each side that the con rod itself meaning that the aluminium piston will rub against phosphor bronze bush not rough (sandblasted remember!) gauge plate.  To ensure that they stuck out the same amount each side a small disk was turned with a 5 thou counterbore and placed on the base of the arbor press then the bush pressed into the counterbore.  The press used was a cheap one, a similar one can be found in Axminster's website on the links page. 

Once the bearing was pressed into position it was time to cut the oil passageway.  The full size rod has an oil slit which breaks through into the bronze bearing.  This is easy enough to replicate either with a woodruff style cutter or small diameter slitting saw, both in the milling machine.  Because you cannot half scale oil i have kept the slit large in width to aide lubrication.  The rod was held in a small toolmakers vice, held in the machine vice on the table.  Due to the small width of the toolmakers vice i was able to clamp onto the parallel section of the rod making holding it very easy.  The slowest speed was selected and the slitting saw was covered in plenty of neat cutting oil. 


The finished items can be seen below.