Hi George,
I had the same anxiety you and Doug have in the "research" phase.
When doing interference press fits you are pretty much committed once
you start ;)
-stop early or keep going...
Removal of AL-AL press fits will cause some spalling? damage and require
a larger insert.
Measure lots, machine under-oversize and finish with 400 grit sandpaper
if you like.
Give yourself some margin based on temperature for the interference fit.
Not hard to do - just time consuming. The epoxy will fill any
imperfections anyway.
Again, practice on a junk housing or whatever to verify your plan will
work and gain confidence.
Decide what you want and Just do it!
Cary
Cary,
I have a lathe but am not a fitter and turner - so If you don't mind I
will
ask some questions.
What difference in sizes did you allow for the interference fit?
What are the expansion (of the housing) and contraction (of the tube) did
you allow for.
I guess what your saying is that you allowed JUST SUFFICIENT to slip fit
the two together, greased with epoxy and when they temp normalized is was
a
good clamped-up fit.
Question, how did you manage to get the tapers in the opposite internal
sides of the water jackets. Tapers on the tube are easier enough.
BTW congratulations - job WELL DONE!!
If you have done it as I assume, this is the optimum sealing method -
there
will be some Engineers on here that would be proud of your effort.
George ( down under)
Hi George;
I am neither a machinist or engineer. Machinists will probably laugh at
this...
I used a basic mill and an abused lathe at our local EAA chapter hanger.
I see you detected my "SUFFICIENT" :) lack of precise measurements
- I still can't find my shop measurements notes, so I'll have to go by
my pickled (beer currently in hand) memory.
IIRC - I did a rough estimate of 220F temp difference (250F rotor, 30F
insert)
x 2.125 hole x 13x10-6 AL in/in/degF coefficient of thermal expansion for
~.006 size increase on the rotor housing - I seem to recall measuring
less than this and had to adjust the insert sizing.
Housing width seemed to track well with temperature change however the
hole size (being unconstrained?) was different.
I think I used a .004-005 total (light) interference fit, with .001-.002
for the slip fit with a mallet and wood block,
pre-greased with epoxy and tapped in. Prototype and first one went in
with minimal effort, 2nd with more effort.
I could not push/pull any insert once inserted and temp equalized. I
did not see any
signs of spalling, the fit appeared good after i wiped away the epoxy
scraped off during the insert process.
I assume the thin wall tube had enough give to deform a little for the
hardest (last) insert.
The insert fit was not perfect, but good enough for aircraft work ;) I
think one (the last insert) bored hole was out of round in the
middle where it was difficult to measure easily. I typically just used
digital calipers (a lot), not the ID telescopic ones (a few times).
BTW- Hot, curing JBWeld fumes smell bad for your health - oh, well, too
late.
I am also unsure of the optimal bond line thickness for JBWeld, some of
the HYSOLs might be an alternative structural
adhesive. Probably doesn't matter for this application anyway.
The rotor housing taper occurred almost naturally with the holesaw cut
and boring head surfacing tool.
The (small) milling machine was at it's limit and this induced vibration
and play. As a result, it naturally cut a slightly conical hole.
Can't remember the taper - maybe .003-.005 over the length of the insert?
I noticed this doing the prototype and decided to take advantage of it.
I match tapered the inserts (+1 screw-up) on the lathe with extra length
just in case. I seem to recall it took very little effort
to misalign the lathe (supported at only one end) correctly.
I hope this method reduces the chances of leaks, but it ain't proven yet.
My intake mounting (flush with the engine face) appears to be a unique
prototype. I could have been less precise
and just pound a tapered insert in till it stuck and added a flange later.
I must reiterate that this method was not difficult but was time
consuming for a one-off - I could do another set much quicker.
I am not sure i would recommend it unless you have the time, patience,
calipers and sandpaper or good CNC machine.
This method is probably OK for one-offs, not for productions runs.
Jerry or one of the CNC guys could say one way of the other.
A larger temperature differential would help with the press fits and the
requirement for accuracy...
For the time I spent, I could have bought a set of MFR cast housings...
As a final note, almost all of the insert PPorts I have heard of were
done using something like Paul's method - almost all seem
to be holding up. Most problems I heard of were failed side port plugs.
It would be hard to draw any conclusions based
on the limited information on problems or failures (since almost no-one
ever discusses problems/failures or details for success...)
Cheers (really)
Cary
Cary,
After my initial request for information, I kinda had this feeling it was a
trial and error type project, due to the lack of specifics, however your on
the right track. I'm sure the process can be refined and catalogued - I'm
working on it.
Thanks for the info!
George ( down under)
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site.
http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.