PDA

View Full Version : Turning an elliptical profile



Michael G
25th Mar 2016, 07:49 PM
Today I intended to make up some gear blanks so I could make the change gears needed for the lathe to cut the screw for the angle plate project but when I pulled the cover off I saw that I'd left it set up to do something else...

One of the guys at work is a car enthusiast or more correctly an engine enthusiast (what funny hobbies some people have - not something sensible like restoring and using fifty year old machine tools:U)
He had been given the patterns and raw casting for some inlet bell mouths but they had been mis-drilled so wanted to know if I could help him get them to a usable condition for his next engine build...
(these sit inside the air plenum and guide the air to get a laminar flow through the carbies. Being for a V8 there were 8 of them - and of course they needed to be more or less identical so that the air flow was all the same)

The castings were pretty ropey (as you will see later) and 3 had been drilled for mounting holes but almost a random hole placement - no marking out or jig drilling here. First order of business was to plug up the drilled holes. The normal way of doing this is to counter sink the hole and then fill with weld. These were Al but the same method applies so out with the TIG and away we went. Ideally you go from one side and then repeat from the other so that there is complete fill but the placement of the holes made that difficult - I had to come back later and re do one casting to get better fill. Al castings can be tricky as the alloys that cast well sometimes don't weld or machine well, but these were alright - makes me wonder what they actually are.
361732
The castings had been linished on the base to make them flat but as this was an air intake ideally the bore of the bell mouth should be square to the flange and the axis of the bore should be the datum that all other features were relative to. Locating on the bore was difficult because the core used to cast was not properly located but eventually I hit on the idea of 2 discs that would locate snugly in the bore. First step was to make that tooling up and a wooden yoke to hold the casting against the disc so that the bottom flange could be faced and the outlet sized.The yoke locates on the two hold down bolts either side of the mount. Clean up as shown in the second photo (your eyes are not playing tricks - the flanging is not even symmetrical)
361731 361730
In that second photo you may just be able to make out the mismatch in the core pattern has not allowed uniform cleanup. Note that if I was planning on doing these again the tooling would be from metal but as it is wood does the job. Once those two features were machined, I could hold the casting onto the form with a special plug and the tailstock.
361729
Because all 8 needed to be the same, I decided to set up to do a cut (the three being face the other side of the flange, turn the OD of the flange, trim the casting to length), locking the respective axis so the distances were the same and swapping through all 8 castings before changing to the next setup. As you can see from the clean up cuts I was doing, these things were just not uniform well made castings! My advice to Paul (the requester) would be to bin these patterns because they produce castings that are all over the shop. I suspect that machining from a piece of bar stock would be just as fast as every surface is going to have to be trued up.
361727 361728
Now the tricky bit - cleaning up the outside. I needed a copy attachment ideally but don't have one so I tried tracing against a 1:1 plot. The basics are shown in the photo below. The white piece of paper held on the tail stock with magnets is a plot of the profile I wanted. I have used a mag base indicator on the compound as a pointer ('cause I could not find anything suitable and could not make anything without breaking set up:doh:) I engaged a slow longitudinal feed and used the cross feed wheel to keep my 'pointer' on the line in the plot. I used a rounded HSS tool so the changing cutting point did not matter, but it worked very well. A bit of polishing with emery and WD40 and they looked the part - certainly better than the sand cast finish they came in
361726
The result so far.
361733
Still to come - repeat the tracing trick on the inside, drill the mounting holes (drill jig already made - it's just visible on the toolbox. Locates in the main bore the holes are in line and symmetrically placed) and trim the edges of the mounting flanges.

Michael

mahgnia
25th Mar 2016, 08:24 PM
Michael,
And you think what I've made sounds like hard work.........:D

Andrew.

Steamwhisperer
26th Mar 2016, 08:50 AM
Ya never cease to amaze me Michael.
Just love picture 7, very clever. :2tsup:

Phil

Michael G
26th Mar 2016, 09:27 AM
Just love picture 7, very clever.

And here's me thinking that method would have been commonly utilised for one offs done before the days of CNC. I was sure that someone like Phil would pop up and say "Yes, we do that all the time when we are reprofiling the hyperbolic steam expansion chambers" (or one of those important but esoteric do-hickies).

Michael

kwijibo99
26th Mar 2016, 12:30 PM
Hi Michael,
I like the improv tracing setup, very clever.
Cheers,
Greg.

Steamwhisperer
27th Mar 2016, 06:42 AM
And here's me thinking that method would have been commonly utilised for one offs done before the days of CNC. I was sure that someone like Phil would pop up and say "Yes, we do that all the time when we are reprofiling the hyperbolic steam expansion chambers" (or one of those important but esoteric do-hickies).

Michael


Love it Michael, not sure of the terminology there but...:D
Not sure when pegboards came into being either but probably a bit later. Not even sure pegboards could do the job.
Of course 'now' I'll have to find out how they did it. We have a 138 year old engine at work with some serious curves on it and I have often wondered...

Phil
ps I still think it was a great bit of lateral thinking:2tsup:

Michael G
27th Mar 2016, 05:33 PM
The inside today - I had hoped that the outside profiles were good enough to line the part up in a 4 jaw chuck but sadly no. I did get to use the 45mm long indicator end to pick up the machined bore at the flange end of the bell mouth though. Another one of those bits of kit that come in the indicator set and you think 'I'll never need to use that!'. The routine then was to centre the bore up in the 4 jaw chuck and bore pretty to the same plan as when doing the OD.
361779
Between setups I did make up a proper pointer which not being as rounded on the end did help. The photo was not all that clear yesterday so here it is again. I found to get clean up I could not follow the profile as drawn but had to do a hybrid arrangement. The second photo is of the chuck board that I made up to protect the ways when changing chucks (basically a board with two lugs to stop it sliding off the ways). Doesn't get used much while changing chucks but damn handy for extra 'putting things down' space. I find with Camlock chucks, holding them/ changing them is not a drama. The previous lathe (where I first made one of these) had screw on chucks - although the chucks were smaller, when screwing them off they would drop when you did not expect it, so the board was used all the time.
361780 361781
Even with the modified profile, there was still a bit of 'as cast' not cleaning up occasionally. Once it moved away from the base, the wall was down to between 4 and 4.5mm on most of these so I was not going to push my luck much more.
361782
Probably 3 of the parts have this sort of issue
361783
Tomorrow all I have to do to these is make the flanges symmetrical around the mounting holes.

Michael

smxas
27th Mar 2016, 10:07 PM
Michael,

I don't know if it is the case on this particular application but I have seen flanges where one side was longer than the other. It was to allow clearance where some inlet port spacing is asymmetrical (centre ones closer together rather than outside ones). I suppose it might not make any difference if they are both shorter, especially since the bell mouths are quite short so don't need additional material for support.

Your friend would probably know from the engine type/design.

Nice work by the way and thanks for posting detailed pictures and description of your work.

Spiro

Michael G
28th Mar 2016, 06:56 AM
It's a point but I gave the drill jig to Paul for him to check for size and he came back and said it fitted, so I'm good. Pardon me if I get this wrong, but something about twin downdraft Hollies was mentioned with one per 2 cylinders (two inlets per unit) - From the photos I was shown they looked symmetrical.
As you say, it's not as if they need lots of material to support. Apart from getting them to look uniform my other concern is not leaving the base too wide in case it obstructs other things. Shouldn't matter that much - after all that work polishing they will sit inside an air plenum, so apart from a smooth bore for flow I guess the rest is not as important.

Michael

jack620
28th Mar 2016, 12:52 PM
The mag pointer and profile sketch is a brilliant idea. One day I intend to make a pulse jet (I really should grow up) but have always pondered how to turn the venturi. Know I know!

Michael G
28th Mar 2016, 09:31 PM
The final bit - more for completeness sake...
I wanted to get the flange symmetric around the mounting holes so thought a jig would be the go. With 4 sides that would be 2 jigs though. Then I had this idea -
361811
The pieces top and bottom are PE, so the Al does not get banged up. However, I have some brass pins in the base piece that lines things up for cutting. Cut one edge then rotate the piece around 180 degrees for the other. Once all 8 parts are done, I slide the pins through and turn the piece around so I can do cuts 3 and 4. All the same angle on the same jig, so can't get more symmetrical than that.
(the brass is not for any special reason - I just had a couple of short lengths lying there so I used them).

Michael