Author Topic: Roy Ozouf's Coventry  (Read 19439 times)

Offline kvom

  • Full Member
  • ****
  • Posts: 1988
Re: Roy Ozouf's Coventry
« Reply #90 on: December 18, 2016, 12:33:08 AM »
After a reassembly with the repaired eccentric strap I had to spend more time and frustration attempting to time it.  Taking a cue from Stan, I made a temporary acrylic steam chest cover (this one really is temporary since I didn't take any care with the overall size.  After more fiddling I got it to run at about 60 RPM (opposite direction this time) but at 80 PSI and leaking air heavily. 

As seen through the "window" the valve looks pretty good, so perhaps I need to try another gasketing material.  I ordered some Loctite 518 and will give that a try when it arrives.

Offline kvom

  • Full Member
  • ****
  • Posts: 1988
Re: Roy Ozouf's Coventry
« Reply #91 on: December 18, 2016, 01:42:03 AM »
After some musings on the valve linkage, I think I have a handle on the proper way to approach timing it.  My thoughts are as follows.

1) The position of the sliding link is solely dependent on the orientation of the eccentric.

2) The cylinder is vertical at TDC and BDC.  That means that if the throw of the eccentric is orthogonal to the crank then the sliding link will be at the same height at both TDC and BDC meaning that the valve will also be at the same position regardless of anything else, and the engine clearly cannot run in that case.  That means in turn that the eccentric throw must be somewhat aligned with the crank.  So the starting point is to align the throw with the crank, with the orientation depending on which direction you want the engine to run.  My early mistakes were to try to adjust the other motions with the eccentric in a randomly installed orientation.

3) Next one can adjust the rocker on its shaft so that it at a minimum doesn't hit the cylinder cover on the downstroke and the setscrew doesn't hit the gland.

4) Now adjust the valve rod in its nut so that the valve has a symmetric range of motion.  To do this I remove the steam chest in order to be able to turn the head of the valve rod.  Then slip the rod end over the rocker and lower the check using its two mounting screws as guides.  Until the eccentric is in final position this is hard to do get exact by eyeball, and later adjustment might be needed.  Since the thread on the valve rod is 3-48, a half turn moves the valve about .01".

5) Now with the crank at TDC adjust the eccentric so that the top edge of the valve is near to opening the top port.  Doing so should slightly reduce the total valve travel so that the same occurs with the bottom edge and port at BDC.  An adjustment of the valve rod might be needed here.  The vertical length of the valve looks critical here as I discovered earlier.

I'll verify whether my theoretical musing match reality in the coming days.

Offline kvom

  • Full Member
  • ****
  • Posts: 1988
Re: Roy Ozouf's Coventry
« Reply #92 on: December 18, 2016, 08:47:42 PM »
Tried the above recipe with mixed success.  I needed a two stage process.  After getting the eccentric  fairly close the range of motion needed to be adjusted once again with the valve rod, and then  fine adjustment of the eccentric.  I also learned that the eccentric adjustment needed to be done with the engine vertical since there remains some slack in the valve linkage at TDC and gravity sucks.

I know we like steam engine models that run at slow RPMs, so here's a really slow one:


Unfortunately that's at 60 psi air pressure.   ::)

Offline Kim

  • Global Moderator
  • Full Member
  • *****
  • Posts: 3636
  • Portland, Oregon, USA
Re: Roy Ozouf's Coventry
« Reply #93 on: December 18, 2016, 07:13:17 PM »
Very nice! And that certainly is really slow!  :popcorn:
Kim

Offline kvom

  • Full Member
  • ****
  • Posts: 1988
Re: Roy Ozouf's Coventry
« Reply #94 on: December 19, 2016, 11:23:03 PM »
As an experiment, I replaced the 1/8" air supply tube with a 1/4" tube.  The increased volume of air meant that the engine can run as low as 25 psi even with its current leakage.  That's encouraging.   :)

Offline MJM460

  • Full Member
  • ****
  • Posts: 856
  • Melbourne, Australia
Re: Roy Ozouf's Coventry
« Reply #95 on: December 20, 2016, 03:06:33 AM »

Hi kvom,

A great thread and a beautiful engine.  Beautifully made.
I am sorry that you you are having such a learning exercise in getting it going, but learning from successes is often limited.  I hope you do not mind if I stick my neck out and offer some suggestions.  I hope it will be helpful, and at worst, we will both learn something if some of our more experienced colleagues are kind enough to step in and correct me.

First, you are correct in noting that with the valve dead centre and the cylinders at top or bottom dead centre, the engine will not run.  That is why a double acting single cylinder engine is not self starting at all positions, but the dead spot is less than a single acting single cylinder engine.  It is the correct position to start your timing.  When the engine is running, the flywheel will take it past this point, and as soon as one inlet port starts to open, another power stroke begins.  Let's assume the design and manufacture of the other side of the valve is correct as we can't see it to check.

Second, very few valve linkages give a truly symmetrical valve motion.  Even the crank and conrod do not give a truly sinusoid all motion.  The main thing is that the valve fully uncovers the steam port, it does not matter if it opens further than this.

The simplest valve exactly spans the outside edges of the steam ports so that starting from piston top dead centre, and the valve mid stroke, any rotation in the right direction starts the power stroke.  You have to work against the steam pressure to rotate it the other way, so the engine will tell you which way it wants to run. 

More advanced design has the valve a little wider than this.  That provides lap and lead which we can discuss later.  Once you know which way it wants to spin, you can rotate the eccentric a little to the point of opening in that direction if you want to.

So first piston top or bottom dead centre, eccentric 90 degrees ahead of the crank in the direction you want to run, adjust the valve position to equally cover both ports, and then adjust the other points of your linkage so it does not interfere anywhere through the complete revolution.

With air pressure and a little spin to start it should run. 

So the next point is to follow through on the valve fit and also fix that leakage.

The valve needs to be a nice free sliding fit on the nut, no slack in the valve rod direction and no friction, like a Stirling engine piston!  However it must have a little slack perpendicular to the valve face so the air or steam pressure can move the valve against the port face well enough to seal.  I assume it goes without saying that the valve face and port face must be flat enough to seal when in contact.  Other tight spots will wear in, or you will help them along with some mechanical procedure, but if the valve nut is tight, the valve will not seal against the port face.  This is one place where a little slack never stopped an engine from running.

You did not mention where the leakage is occurring, or did I miss it, but you did mention gaskets.  So I assume not valve blow by, but a leaky joint.  Liquid gaskets are really good when you are on your final assembly, but tend to be difficult if you still have a few strip downs for final tuning.  I use brown paper gaskets, preferably not a highly glazed type, but one that will absorb a bit of oil.  Smear both sides of the gasket with light oil, and assemble tightening bolts evenly.

I am sure that you know most of this stuff, but I hope that I might just have mentioned something critical that you have overlooked.

MJM
The more I learn, the more I find that I still have to learn!

Offline kvom

  • Full Member
  • ****
  • Posts: 1988
Re: Roy Ozouf's Coventry
« Reply #96 on: December 20, 2016, 04:39:58 AM »
That's a very good summary.  I have build a number of slide valve engines where the valve rod is attached to the eccentric is a fairly rigid manner so that there's no appreciable slack.  The problem with this one is that there are 7 separate parts from the eccentric to valve inclusive, meaning any fitment errors can add up more than we like.

In this engine the valve is 1/32" wider than the ports, so there is that much potential lead.

For the steam chest I introduced an additional part:  a  valve plate made from ground steel sheet primarily to assure a good sliding surface for the valve.  So the chest is a 4 layer sandwich meaning 4 surface mates subject to leaking.  It's all held together with 2-56 screws, so I can't really clamp down too hard.  I can feel air escaping to one side of the chest, and when I press the layers together manually the leak lessens.  Next session I want to take the chest apart, clean the parts, and examine the fits more closely.


Offline kvom

  • Full Member
  • ****
  • Posts: 1988
Re: Roy Ozouf's Coventry
« Reply #97 on: December 20, 2016, 11:06:12 PM »
Disassembled and removed all the gasket material by soaking in acetone.  I'll try the loctite gaskets next;  it should arrive in the mail tomorrow or Thursday.

Offline kvom

  • Full Member
  • ****
  • Posts: 1988
Re: Roy Ozouf's Coventry
« Reply #98 on: January 22, 2017, 04:41:47 PM »
A few posts were possibly lost in this thread, but I did take the engine to Cabin Fever, where it ran slowly at times on the 40 psi air provided at the display table.  It needed its own spigot on the air pipe as running other engines on the same manifold reduced the available flow too much.

Got it home and disassembled the lot, cleaned the port block and cylinder of old sealant, and came up with a new plan.  I went back to the Permatex aircraft gasket sealant, but figured I needed to press the port block into the cylinder rather than slide it in in order to keep from scraping the sealant off the mating surfaces.  So my technique was as follows.

First, apply to the bottom of the port block fairly thinly, and wipe away any sealant in the supply and exhaust channels.  Leave exposed to air for several minutes as recommended by Permatex.  Next, insert 6 screws into the tapped holes in the cylinder to keep sealant out of the holes.  Here I brushed on a thicker layer than on the port block.  After letting it sit for several minutes, I removed the screws and used a paper towel to soak up any sealant around the holes.  Next I inserted 6 long screws though the pot block and used them to screw into the cylinder.  These screws would then guide the port block as it was pressed into the cylinder.  I did the press using the vise and a piece of 1" brass rod against the port block.  Finally wiped away any sealant that squeezed out.  I removed the 6 long screws and replaced them with the 4 short screws per plan. however I'm pretty sure that they're unnecessary given the tight fit of the block.

After several hours of curing, I screwed the steam chest cover onto the port block attached the air to the pivot shaft; I was happy to not feel any leaks around either the port block or the cylinder covers.  After removing the steam check cover I applied air to the steam port openings in the port block and saw that the piston moved in both directions, confirming that the sealant did not cause any blockage.

After reassembling it was the moment of truth.  Engine started running immediately at 20 PSI, and continued to do so at 10.  In this video it's running at 10, and then I increase it to 20.


Offline vcutajar

  • Full Member
  • ****
  • Posts: 2475
  • Marsascala, MALTA
Re: Roy Ozouf's Coventry
« Reply #99 on: January 22, 2017, 04:52:49 PM »
Nice one KVOM.

Vince

Offline sshire

  • Full Member
  • *****
  • Posts: 3044
    • LS Editions
Re: Roy Ozouf's Coventry
« Reply #100 on: January 22, 2017, 05:05:49 PM »
Oooooh! I have the aircraft sealant but was waiting for your results.
I'm on it.
Thanks
Best,
Stan

Offline b.lindsey

  • Global Moderator
  • Full Member
  • *****
  • Posts: 13591
  • Dallas, NC, USA
    • Workbench-Miniatures
Re: Roy Ozouf's Coventry
« Reply #101 on: January 22, 2017, 06:16:06 PM »
From 60 psi down to 10 psi...obviously the fine tuning was a great success. And the smoothness of seeing it run proves it!!  Well done and a lovely model  :ThumbsUp:

Bill

Offline kvom

  • Full Member
  • ****
  • Posts: 1988
Re: Roy Ozouf's Coventry
« Reply #102 on: February 20, 2017, 11:25:51 PM »
I finally got the mini-CNC lathe working "reasonably", so it was time to make the decorative turning on the frame spreaders.

Here's the setup:



The screw on the chuck end was necessary to allow enough room for the tool.  Finish is not great and there was chatter on this one.  Played around with feeds, and the last one came out the best.  Some work with emery paper and scotchbrite made them look a bit better.

I am not that happy with the inserts on the Shars turning tools.  Finish in brass is just OK, and it terrible on aluminum.  I ground a HSS bit to try on AL, and that worked much better.  Still I prefer inserts, so will likely try the Arthur Ward HSS inserts.

Installed on the engine:


Offline gary.a.ayres

  • Full Member
  • ****
  • Posts: 479
  • British Isles & sometimes France
Re: Roy Ozouf's Coventry
« Reply #103 on: March 14, 2019, 09:38:33 PM »
Very interesting build log and a beautiful, well-made engine.

Way beyond my skill level but I can appreciate anyway...

Offline Johnmcc69

  • Full Member
  • ****
  • Posts: 479
  • Erie Pa., USA
Re: Roy Ozouf's Coventry
« Reply #104 on: March 15, 2019, 12:16:51 AM »
Nice work Kvom!
 I like the design of the "stays" at the top of the supports, I saw George B. Use the same technique & will copy it on one of the next designs I have in the works.

 Some really nice work you've done,

 John