Home‎ > ‎Part Index‎ > ‎

Engine Cover

A website dedicated to the construction of an accurate 1/2 scale replica of a 1937 Aero Douglas Motorcycle
The timing case or engine cover is quite a complex casting having complex 3D forms but also having a constant thickness of slightly over 1/8".  The half scale cover was machined from the solid using the X3 CNC mill.  Once again the first stage was to create a 3D model of this complex shape using SolidEdge V20 software.

There are many bulges on the front, including the 'nose' which will support the oil sight feed and the raised circular area where the dynamo will be bolted to.  It took many hours to get the 3D model to look like the full size  one.  The biggest difficulty encoutnered was the height of the nose as this would require a cutter with long enough flutes to cut the full depth but be small enough in diameter to recreate the smaller details and get in the gaps such as between the bottom of the nose and the raised lip of the dyanamo bulge. 

A 1/4" ball nose cutter was found in the J&L industrial supply catalogue which had a flute length of 1.5" which was suitable, two were purchased and designated roughing and finishing.  The 3D model also included the rear detail which is almost as complicated as the front, having as previously said, constant thickness with the front plus oil passageways and an area for the oil pump to bolt to.  By doing the complete 3D model I was able to see of any problems ahead in manufacture such as holes breaking through etc.  I then saved the 3D model as a stl file which is the standard transfer file for Vectrics Cut3D program and other general CNC/Rapid Prototyping programs.  A stl viewer was then used to finally inspect the model.  A really good function of this program is you can slice the model at any point controlled by a slider, meaning you can see the entire form of a model in the X, Y or Z in any position allowing checks of thickness etc. to be carried out.  The link to the freeware program used is in the links page. 
The program was written using Vectrics Cut3D as on the Crankcase.  




Once the model and program was written and I was totally happy with it, an aluminium blank was found.  The aluminium was machined on the manual mill to remove a large quantity of the excess material that would only be turned to swarf and blunt the cutter in the process.  Even though the cutter used was a tungsten carbide cutter - it was also expensive and I want it to last as long as possible.  The roughed out blank can be seen sat next to its full size counterpart - the case seen here is from my 'spare' engine.  The accurate work now begins with the rear of the blank having the bolt holes drilled and tapped.  These holes were tapped a smaller size than the hole required when finished.  This allowed the blank to be accuratly bolted down using these holes then when the cover is finished, they can be drilled through clearence for the bolts used into the crankcase.  The holes were not drilled through all the way but as the blank is machined they will be revealed. 

The blank is bolted down to a raised aluminium plate that was machined flat and gave access to underneath so that cap-head bolts could be used to hold the blank down.  The matching holes in the plate were drilled with very little to no clearence on the bolts to ensure maximum location accuracy and then a datum point on the plate was found so that it could be used many times after the machine had be restarted as per the crankcase.  All this has already been done as the plate and bolting system was used for the crankcase and was not removed in the meanwhile.  A regular supply of cutting oil was used and the machine made ready for the mamoth operation.  The cutting time did not take quite as long as the crankcase but it still took enough time that it had to be stopped and restarted the next day.     
 
Photograph 3 shows the roughed out blank of the front cover with several features emerging from the solid.  This only took a 'working day' to produce as although the steps in the X were only 20 thou per pass, the depth was 60 thou per pass with approx 10 thou left on for a final pass.  Of course with such a rough depth on the roughing out, alot of the detail has not come out, leaving steps which are quite visible.  This means that there is allot more material to be taken off when the final pass was taken.  As the final pass is quite slow and it is a very rigid setup it took the extra depth cuts without trouble.  The datum used can be seen in the top right corner of the picture, the top of this was also used for the Z datum as the machine was reset between roughing and finshing operations.  

Photograph 4 shows the final cut complete on the front of the blank.  There is alot of cleaning up as the cutter chattered fairly badly on the deep vertical surfaces, which with a 1/4" diameter cutter with over 1.5" cutting is to be expected.  Also there is ALOT of polishing to be done to remove the machining marks.  All the now exposed bolting down holes were then accuratly spot faced to the same height which was important for the next process. 



The rear of the blank was then flycut to thickness and in doing so, chopped off the unwanted corners left on by the front process.  As you can see the blank was required to be in the same position as when the front was cut but this was difficult due to the large nose sticking out, so 10 posts with accurate lengths, threaded one end and tapped the other were screwed into the blank then bolted to the plate similar to the previous operation.  This formed a very sturdy frame and coupled with light cuts proved strong enough for the machining of the rear. 

A similar process was used for the rear of the cover which is just as complex due to the oil passage ways and the original cover being constant thickness.  Then a needle file and some fine emery paper was used to polish out the criss-cross pattern and any chatter given by the program. 

There is an Oil Feed window to be bolted to the 'nose' and a 'BTH' pancake dynamo to be bolted onto the raised circular area.  On the rear an oil pump is to be bolted to the large flat run via a gear pivoting around an axle located on the central raised spigot.