Terry Mayhugh's fabricated exhaust manifolds

Here are some pictures of the exhaust manifolds that I just completed for my V-4. I machined this from billet and then glass-beaded them to give them a 'cast' look. In another thread I showed how I created the intake manifolds from two identical halves, but I felt the exhaust manifolds needed to be machined from a single workpiece due to the extreme temperatures to which they would be exposed. I started with a suitably sized rectangular workpiece and then it it I drilled the long exhaust runner. I then drilled the two short runners as far as I could before the angle into the block flange became too acute. I then pressed plugs into the short runners to a depth such that half of the diameter of the long runner would be closed off. I also pressed pins into these plugs where the wall the manifold itself would eventually be as a safety measure to help secure the plug. Plugs and pins were also secured with high temperature bearing retainer. I then re-drilled the long runner to clear out the remnants of the two short runner plugs. I used a spherical D-bit to blend the long runner with the far-end short runner. I then machined the the top of the manifold and then flipped it over to mill the backside. This is where things get tricky. Finish machining the backside will cause the part to drop free before the machining is completed and so before starting the backside machining I epoxied the already top-side machined part to a plate that was secured to the workpiece with a number of screws. This kept the part stable after it was cut free from the workpiece and allowed me to do all the back side machining. After both sides were completed I heated the epoxy up to 200F and it released easily from the workpiece. Essentially no clean-up was needed and so I needed only to drill out the short runners from the block flanges to connect the runners to the already drilled portions. I then bead-blasted the parts to make them look cast. Two months of development work and about 4 hours of machining time per manifold. I'm glad its over but I'm happy with the result. -Terry