Williams ws. Rankin
Posted by Howard Langdon
|
Re: Williams ws. Rankin October 03, 2006 11:06AM |
Registered: 19 years ago Posts: 1,038 |
Hi Harry
The S.E.S. engine should not have had the breathing problems like the Stanley, White or Doble. The Doble had better breathing then the older White or Stanley. Acording to Jim. <[www.steamautomobile.com]; <[www.steamautomobile.com];
The old piston and slide valve engines had long ports and would be quite restrictive. The Doble was an improvement having larger ports. But the S.E.S. engin had a farly stright shot through the valves. S.E.S. were especially conserned with good breathing.
So we we have a history of improved breathing. Yet the torque drop off at around 900 RPM existed in all thoes engines.
You can believe in high RPM but belief doesn't make anything a fact. What I wont is scientific proof that the charistric torque drop off at 900 RPM has been vanquished. I would love to go to higher RPM to have a smaller engine. Running at high RPM with no load means very little. I have played a bit with small engines and they can atain close to 3000 RPM with no load. But with the slightest load they drop to 1000 RPM.
All I am asking for is a simple dyno run. What exactly does yor engines torque curve look like?
Andy
The S.E.S. engine should not have had the breathing problems like the Stanley, White or Doble. The Doble had better breathing then the older White or Stanley. Acording to Jim. <[www.steamautomobile.com]; <[www.steamautomobile.com];
The old piston and slide valve engines had long ports and would be quite restrictive. The Doble was an improvement having larger ports. But the S.E.S. engin had a farly stright shot through the valves. S.E.S. were especially conserned with good breathing.
So we we have a history of improved breathing. Yet the torque drop off at around 900 RPM existed in all thoes engines.
You can believe in high RPM but belief doesn't make anything a fact. What I wont is scientific proof that the charistric torque drop off at 900 RPM has been vanquished. I would love to go to higher RPM to have a smaller engine. Running at high RPM with no load means very little. I have played a bit with small engines and they can atain close to 3000 RPM with no load. But with the slightest load they drop to 1000 RPM.
All I am asking for is a simple dyno run. What exactly does yor engines torque curve look like?
Andy
|
Re: Williams ws. Rankin October 03, 2006 11:22AM |
Registered: 19 years ago Posts: 1,268 |
|
Re: Williams ws. Rankin October 03, 2006 12:41PM |
Registered: 19 years ago Posts: 1,038 |
|
Re: Williams ws. Rankin October 03, 2006 02:16PM |
Registered: 19 years ago Posts: 835 |
Hi Andy,
Jay Carters engine was 90 hp at 5000rpm that is 94.5ft lbs torque. I have built a lot of other engines. Have you built any using a bump valve? Bill Ryan's
go cart is an example of a high rpm engine delivering a lot of power in a small package at high rpm. The model boats are also an example, mind you they are small and it is piston speed that counts.
There are two of us designing and building all of this equipment. There is over 100 drawings on the 6cyl engine alone. We are building and testing these parts as the engine is the simple part as it is basicly the sum of its parts. We were testing pumps last week this week testing the combustion chamber burning orange peal juice. now we can turn on the fire like a light switch with a clean burn. As I told you before the info on dyno testing as well as fuel curves, air quality ect. will be presented in a paper. You are welcome as well as other SACA members to visit, as a matter of fact one is coming tomorrow. Others have been here. Please come.
Harry
Jay Carters engine was 90 hp at 5000rpm that is 94.5ft lbs torque. I have built a lot of other engines. Have you built any using a bump valve? Bill Ryan's
go cart is an example of a high rpm engine delivering a lot of power in a small package at high rpm. The model boats are also an example, mind you they are small and it is piston speed that counts.
There are two of us designing and building all of this equipment. There is over 100 drawings on the 6cyl engine alone. We are building and testing these parts as the engine is the simple part as it is basicly the sum of its parts. We were testing pumps last week this week testing the combustion chamber burning orange peal juice. now we can turn on the fire like a light switch with a clean burn. As I told you before the info on dyno testing as well as fuel curves, air quality ect. will be presented in a paper. You are welcome as well as other SACA members to visit, as a matter of fact one is coming tomorrow. Others have been here. Please come.
Harry
|
Re: Williams ws. Rankin October 03, 2006 07:23PM |
Registered: 19 years ago Posts: 318 |
|
Re: Williams ws. Rankin October 04, 2006 10:09AM |
Registered: 19 years ago Posts: 1,441 |
Andy, Howard, et al,
There is only one way to find out why torque diminishes at this 900 rpm supposed limit. Or any other problem with a new engine design.
Put the suspect engine on a calibrated dyno. Then install fast response transducers in the heads each of the cylinders. Then a sensor on the shaft to trigger the scope for horizontal sweep. This is exactly how and why Besler developed his engines the way he did, and the resulting diagrams showed PRECISELY what was going on in the cylinder at various loads and speeds. All conditions are then clearly revealed and corrections could be made from hard data.
Now you will SEE in the PV diagram on the scope, just what is going on and when and why. Anything else is simply fooling around and making guesses.
This is why I always have maintained that building a new steam car engine is only one step in the process, developing it is where the big money is going to be spent. Anything less is simply playing around in the parking lot.
Designing and making and developing a new steam car engine is one very expensive project. Harry, I think you have seen this with your engine.
Howard, the Doble F engine was designed to use 850° steam at 1500 psi; but they had to back down on this because of the oil breaking down and making carbon, and fast ring wear.
JC
There is only one way to find out why torque diminishes at this 900 rpm supposed limit. Or any other problem with a new engine design.
Put the suspect engine on a calibrated dyno. Then install fast response transducers in the heads each of the cylinders. Then a sensor on the shaft to trigger the scope for horizontal sweep. This is exactly how and why Besler developed his engines the way he did, and the resulting diagrams showed PRECISELY what was going on in the cylinder at various loads and speeds. All conditions are then clearly revealed and corrections could be made from hard data.
Now you will SEE in the PV diagram on the scope, just what is going on and when and why. Anything else is simply fooling around and making guesses.
This is why I always have maintained that building a new steam car engine is only one step in the process, developing it is where the big money is going to be spent. Anything less is simply playing around in the parking lot.
Designing and making and developing a new steam car engine is one very expensive project. Harry, I think you have seen this with your engine.
Howard, the Doble F engine was designed to use 850° steam at 1500 psi; but they had to back down on this because of the oil breaking down and making carbon, and fast ring wear.
JC
|
Re: Williams ws. Rankin October 04, 2006 10:42AM |
Registered: 19 years ago Posts: 835 |
Hi Jim
You said it much better than I did. We ran engines over a year and half ago and there is so much to learn. You start to learn when you lite a fire. Any thing with a piston and a crank will go around. When you burn any fuel clean and condence completely in a small package that is efficent you are just starting.
Harry
You said it much better than I did. We ran engines over a year and half ago and there is so much to learn. You start to learn when you lite a fire. Any thing with a piston and a crank will go around. When you burn any fuel clean and condence completely in a small package that is efficent you are just starting.
Harry
|
Re: Williams ws. Rankin October 04, 2006 01:37PM |
Registered: 19 years ago Posts: 1,038 |
Hi Jim.
Regrding the 900 RPM torque drop off.
What you said is exactly my point. Harry is caliming to have solved the problem. But no hard proof has been shown.
It is only speculation that flow is the problem. We don't have any data or tests to say what the reasion for the torque drop of is.
I would say the flow restriction is the most probable cause, except for test data from several engines that should have quite different flows and they all have that same 900 RPM torque drop off. The drop off rates are a little different. But the start of the drop off is almost identical.
I can't rule out flow. But with the dyno data available it seams unlikely.
That is why I am after Harry to produce some dyno data. If he has solved the problem a dyno run would prove it, but wouldn't tell us the reasion. Harry and Jay both use very high pressures. Perhaps it's the density that does it. But at this point we don't know if thoes high RPM engines of Jay's amd Harry's do solve the drop off problem.
Anyway! Doesn't look like Harry is going to release the data any time soon.
Andy
Regrding the 900 RPM torque drop off.
What you said is exactly my point. Harry is caliming to have solved the problem. But no hard proof has been shown.
It is only speculation that flow is the problem. We don't have any data or tests to say what the reasion for the torque drop of is.
I would say the flow restriction is the most probable cause, except for test data from several engines that should have quite different flows and they all have that same 900 RPM torque drop off. The drop off rates are a little different. But the start of the drop off is almost identical.
I can't rule out flow. But with the dyno data available it seams unlikely.
That is why I am after Harry to produce some dyno data. If he has solved the problem a dyno run would prove it, but wouldn't tell us the reasion. Harry and Jay both use very high pressures. Perhaps it's the density that does it. But at this point we don't know if thoes high RPM engines of Jay's amd Harry's do solve the drop off problem.
Anyway! Doesn't look like Harry is going to release the data any time soon.
Andy
|
Re: Williams ws. Rankin October 04, 2006 09:05PM |
Registered: 19 years ago Posts: 318 |
|
Re: Williams ws. Rankin October 05, 2006 12:57AM |
Registered: 19 years ago Posts: 1,268 |
|
Re: Williams ws. Rankin October 05, 2006 10:40AM |
Registered: 19 years ago Posts: 1,441 |
Andy,
No, Harry is not going to release any data that he cannot firmly support with hard factual numbers. This approach is PRECISELY the right way to go, good for you Harry.
A learned paper would certainly be my course of action too, when ALL numbers are verified.
Just what engine had this 900 rpm limit, and how was that figure established?
Also, did "they" give any reasons for this?
Bill,
Photos today. Yesterday at lunch the "victim" vehicle just may have been secured, the Series III V-12 Jag roadster. Now the question, do I tidy up the car, fix the leaky main seal, paint it, and enjoy it, or start making a fool out of myself and start one more steam car? Something I promised myself never to do again; but the bug certainly has bitten years ago and very hard.
JC
No, Harry is not going to release any data that he cannot firmly support with hard factual numbers. This approach is PRECISELY the right way to go, good for you Harry.
A learned paper would certainly be my course of action too, when ALL numbers are verified.
Just what engine had this 900 rpm limit, and how was that figure established?
Also, did "they" give any reasons for this?
Bill,
Photos today. Yesterday at lunch the "victim" vehicle just may have been secured, the Series III V-12 Jag roadster. Now the question, do I tidy up the car, fix the leaky main seal, paint it, and enjoy it, or start making a fool out of myself and start one more steam car? Something I promised myself never to do again; but the bug certainly has bitten years ago and very hard.
JC
|
Re: Williams ws. Rankin October 05, 2006 11:28AM |
Admin Registered: 18 years ago Posts: 2,107 |
I have to just note that the much maligned GM SE-101 4 cylinder uniflow expander had an output of 160 HP at 3,200 rpm. The cylinders were oversquare at 3.5 by 2.624 and the overhead poppet valve was a proportionally generous 1.5 inch diameter by .102 inch lift. Cutoff was about 30 percent but the gentle cam profile meant that the majority of the closure occurred much earlier, examination of expansion profiles indicated an effective cutoff of about 15%. I'm presuming the relatively generous valve size and unrestricted flow path made the engine free breathing and fairly quick.
I long ago noted a design by Fitchburg, an old line industrial steam engine manufacturer, that looked like a promising starting point for uniflow engines. See attachment.
Regards,
Ken
I long ago noted a design by Fitchburg, an old line industrial steam engine manufacturer, that looked like a promising starting point for uniflow engines. See attachment.
Regards,
Ken
|
Re: Williams ws. Rankin October 05, 2006 11:41AM |
Registered: 19 years ago Posts: 375 |
|
Re: Williams ws. Rankin October 05, 2006 12:22PM |
Registered: 19 years ago Posts: 1,038 |
Hi Jim & Howard.
The torque data for Doble came from the Doble book, borrowed, don't remember the exact title or author. It's been metioned here before. It had a torque chart for one of th engines. Didn't describe in detail how it was done. Wether fixed cutoff or if cutoff were changed. The whate data came from a paper on White performance by some perfessor. I probably still have the paper some where. The torque plot for the white was done on a dyno at short cutoff. So it was a fixed cutoff for the white. The S.E.S. data is fairly detailed. It was using a fixed short cutoff also. I know that the white and S.E.S. data were for a fixed cutoff. The S.E.S. data came from a HP chart I converted to torque. It wasn't a self starter. But it went to low enough RPM that one could easly see the trend that max torque would have been close to 0. From the lowest RPM to 850 RPM these engine have a flat torque line. I put the date into a spread sheet. Scaling the torque so the max coiensided the plots lay nearly on top of each other.
Howard, The torque data is from books and papers on thoes engines. The S.E.S. paper is the most detailed It's a fairly short cutoff. Will have to find the paper to say for certain but I think its like between 4 to 10% cutoff. I'm faily sure it was a uniflow engine. The White was at a short cutoff notch. Don't remember which one. Will have to find the paper on that one. There wasn't much detail on the Doble though. It stated a specific engine if I remember corectly.
The White and S.E.S. runs were at a fixed cutoff. Don't know about the Doble. But sense it's almosy exactly the same shape I would think it to be the same as the others.
Andy
The torque data for Doble came from the Doble book, borrowed, don't remember the exact title or author. It's been metioned here before. It had a torque chart for one of th engines. Didn't describe in detail how it was done. Wether fixed cutoff or if cutoff were changed. The whate data came from a paper on White performance by some perfessor. I probably still have the paper some where. The torque plot for the white was done on a dyno at short cutoff. So it was a fixed cutoff for the white. The S.E.S. data is fairly detailed. It was using a fixed short cutoff also. I know that the white and S.E.S. data were for a fixed cutoff. The S.E.S. data came from a HP chart I converted to torque. It wasn't a self starter. But it went to low enough RPM that one could easly see the trend that max torque would have been close to 0. From the lowest RPM to 850 RPM these engine have a flat torque line. I put the date into a spread sheet. Scaling the torque so the max coiensided the plots lay nearly on top of each other.
Howard, The torque data is from books and papers on thoes engines. The S.E.S. paper is the most detailed It's a fairly short cutoff. Will have to find the paper to say for certain but I think its like between 4 to 10% cutoff. I'm faily sure it was a uniflow engine. The White was at a short cutoff notch. Don't remember which one. Will have to find the paper on that one. There wasn't much detail on the Doble though. It stated a specific engine if I remember corectly.
The White and S.E.S. runs were at a fixed cutoff. Don't know about the Doble. But sense it's almosy exactly the same shape I would think it to be the same as the others.
Andy
|
Re: Williams ws. Rankin October 05, 2006 12:35PM |
Admin Registered: 18 years ago Posts: 2,107 |
Hi Mike:
Easiest way to find out more would be to go to <www.pat2pdf.org> and type in the patent number, in this case 2,113,936. The website will look up the patent in the USPTO files and convert it into an Adobe .pdf file that you can download. I find this much easier than dealing with the patent office website if I already know the patent number.
The engine has a hydraulic valve gear and as you noted a spring in the steam. With Inconel or something similar you could likely live with the spring. Given my own drathers I'd dump the hydraulically actuated valve system (which probably looked cool and futuristic in the 30's) and install a camshaft and rocker arm system with hydraulic lifters, external spring and possibly a cam phaser to adjust cam angle at speed.
Regards,
Ken
Easiest way to find out more would be to go to <www.pat2pdf.org> and type in the patent number, in this case 2,113,936. The website will look up the patent in the USPTO files and convert it into an Adobe .pdf file that you can download. I find this much easier than dealing with the patent office website if I already know the patent number.
The engine has a hydraulic valve gear and as you noted a spring in the steam. With Inconel or something similar you could likely live with the spring. Given my own drathers I'd dump the hydraulically actuated valve system (which probably looked cool and futuristic in the 30's) and install a camshaft and rocker arm system with hydraulic lifters, external spring and possibly a cam phaser to adjust cam angle at speed.
Regards,
Ken
|
Re: Williams ws. Rankin October 05, 2006 01:36PM |
Registered: 19 years ago Posts: 1,038 |
|
Re: Williams ws. Rankin October 05, 2006 03:34PM |
Registered: 19 years ago Posts: 375 |
Ken,
Nice way of looking at patents, thanks.
The main claim is the hydraulic lifter which looks pretty but makes me wonder if the thing was actually built.
The valve opens against the steam pressure as usual but because of the set up the opening action of the lifter is to pull on the valve steam not push. This would be hard to do mechanically, may be this is the reason for the hydraulics.
The snag is that the valve stem seal is exposed to full steam pressure all the time, not just to the reduced steam pressure in the port during the open phase.
Andy,
The secong ring set below the exhaust port - presumably as these rings are only coping with exhaust steam pressure they would do the job well enough.
Mike
Nice way of looking at patents, thanks.
The main claim is the hydraulic lifter which looks pretty but makes me wonder if the thing was actually built.
The valve opens against the steam pressure as usual but because of the set up the opening action of the lifter is to pull on the valve steam not push. This would be hard to do mechanically, may be this is the reason for the hydraulics.
The snag is that the valve stem seal is exposed to full steam pressure all the time, not just to the reduced steam pressure in the port during the open phase.
Andy,
The secong ring set below the exhaust port - presumably as these rings are only coping with exhaust steam pressure they would do the job well enough.
Mike
|
Re: Williams ws. Rankin October 05, 2006 04:08PM |
Registered: 19 years ago Posts: 1,441 |
Bill,
This forum is not letting me put more than one of the pictures on a response.
Will try sending them to you on your e-mail address.
Just tried that and that doesn't work either. Got the forum private message thing.
Send me your e-mail address.
jdcrank@pacbell.net
JC
Edited 1 time(s). Last edit at 10/05/2006 04:13PM by James D. Crank.
This forum is not letting me put more than one of the pictures on a response.
Will try sending them to you on your e-mail address.
Just tried that and that doesn't work either. Got the forum private message thing.
Send me your e-mail address.
jdcrank@pacbell.net
JC
Edited 1 time(s). Last edit at 10/05/2006 04:13PM by James D. Crank.
|
Re: Williams ws. Rankin October 05, 2006 04:18PM |
Registered: 19 years ago Posts: 1,038 |
Jim,
Havn't had problems ataching to posts. But thay must be an accepted type of document. JPG PDF etc. Check with Scott for types. Don't know what format your trying to attach but there are ways to convert from one to another. Adobe photoshop handles all kinds. PaintShop Pro is another good program. You can also find freebees that will work fine. Generally with the photo and paint programs you load the image and do a "Save As" to select a diferent save format.
Andy
Havn't had problems ataching to posts. But thay must be an accepted type of document. JPG PDF etc. Check with Scott for types. Don't know what format your trying to attach but there are ways to convert from one to another. Adobe photoshop handles all kinds. PaintShop Pro is another good program. You can also find freebees that will work fine. Generally with the photo and paint programs you load the image and do a "Save As" to select a diferent save format.
Andy
|
Re: Williams ws. Rankin October 05, 2006 06:54PM |
Admin Registered: 18 years ago Posts: 2,107 |
Hi Mike:
I think you'll find the energy needed to lift the poppet is much less than you anticipate. This is a uniflow engine, so there is recompression of remnant steam left in the cylinder. Typically the designer hopes to recompress the steam to something close to the admission pressure. This would roughly balance the poppet valve, leaving fairly low opening forces.
I also note that they put a labyrinth seal on the valve stem, each step in the stem causes a pressure drop and reduces leakage, worked pretty well on main propulsion turbines I worked with in the navy. Personally, I'd consider putting a packing ring or two on the poppet stem and as I noted earlier, dump the hydraulic mechanism for a fairly conventional current day valve train. To be fair to the designer, at the time that engine was patented such hydraulic mechanisms were all the rage in design circles, you can see it in a lot of patents from the period. Usually you come back to the classics, though. Of course, if you design the hydraulics right, you can have higher oil pressure than steam pressure and have a slight hydraulic fluid leak along the valve stem into the cylinder and handle your cylinder lube that way. Just thinking out loud here for my super duper, super secret, overhead valve injection cycle engine....
Andy: I read that the Project 77 steam VW conversion used similar rings on the bottom of a long uniflow piston. The published reports indicate that about one teaspoon of water crept in after some hours of operation. I think if they had subatmospheric condensing maybe even less would get in as the flow would be past the piston to the exhaust ports. Of course, then air would be creeping up through the crankcase into your condenser. Personally, I'd consider taking a page from old Abner's book; he planned to run the Ultimax crankcase under vacuum and employ a good mechanical seal on the crankshaft. Mechanical seals run for thousands and thousands of hours without fail, and aren't that expensive so this looks valid to me. At subatmospheric pressure and typical oil temperatures, the residual water should vaporize out, especially if a good synthetic oil was used.
Regards,
Ken
I think you'll find the energy needed to lift the poppet is much less than you anticipate. This is a uniflow engine, so there is recompression of remnant steam left in the cylinder. Typically the designer hopes to recompress the steam to something close to the admission pressure. This would roughly balance the poppet valve, leaving fairly low opening forces.
I also note that they put a labyrinth seal on the valve stem, each step in the stem causes a pressure drop and reduces leakage, worked pretty well on main propulsion turbines I worked with in the navy. Personally, I'd consider putting a packing ring or two on the poppet stem and as I noted earlier, dump the hydraulic mechanism for a fairly conventional current day valve train. To be fair to the designer, at the time that engine was patented such hydraulic mechanisms were all the rage in design circles, you can see it in a lot of patents from the period. Usually you come back to the classics, though. Of course, if you design the hydraulics right, you can have higher oil pressure than steam pressure and have a slight hydraulic fluid leak along the valve stem into the cylinder and handle your cylinder lube that way. Just thinking out loud here for my super duper, super secret, overhead valve injection cycle engine....
Andy: I read that the Project 77 steam VW conversion used similar rings on the bottom of a long uniflow piston. The published reports indicate that about one teaspoon of water crept in after some hours of operation. I think if they had subatmospheric condensing maybe even less would get in as the flow would be past the piston to the exhaust ports. Of course, then air would be creeping up through the crankcase into your condenser. Personally, I'd consider taking a page from old Abner's book; he planned to run the Ultimax crankcase under vacuum and employ a good mechanical seal on the crankshaft. Mechanical seals run for thousands and thousands of hours without fail, and aren't that expensive so this looks valid to me. At subatmospheric pressure and typical oil temperatures, the residual water should vaporize out, especially if a good synthetic oil was used.
Regards,
Ken
|
Re: Williams ws. Rankin October 05, 2006 06:58PM |
Admin Registered: 18 years ago Posts: 2,107 |
Hi Jim:
I had to try two times earlier today before the forum software accepted my attachment, the previous two tries it not only wouldn't take the attachment but dumped my message as well. Apparently some kind of issue with posting attachments as I did exactly the same process all three times.
Regards,
Ken
I had to try two times earlier today before the forum software accepted my attachment, the previous two tries it not only wouldn't take the attachment but dumped my message as well. Apparently some kind of issue with posting attachments as I did exactly the same process all three times.
Regards,
Ken
|
Re: Williams ws. Rankin October 06, 2006 12:50AM |
Registered: 19 years ago Posts: 1,268 |
Hello,
Looking at what I can get from the data on the GM engine then the cylinder 3.5" dia by 2.6" stroke is 25" cubed. The valve at 1.5" dia is a circumference of 4.7" and with a lift of 0.102" then has an inlet port area of 0.48".
This is then at 15% cutoff a volume of 3.8" roughly of new steam.
So 0.48/3.8 = .12 as a valve area to charge volume factor.
Mine has a valve port area of 0.6 to a charge volume at 15% cutoff of 2.4" cubed.
This yeilds a factor of .25 as a valve area to charge volume factor.
It's a lousy comparison method as other factors should enter in, but it is something. Apples to apples it would be twice the valve area.
The thing to notice is that the Gm engine was/is considered a more modern one yet still took over 20 degrees to finish closing the valve. One which didn't open that far to begin with. With boundary layers of up to 20 thou depending on steam velocitys an opening of 0.102 starts to choke quick. The Williams valves of around 0.050 max lift are almost choked when full open.
These low lifts are expensive when considering that the constant pressure process is not happening at inlet pressure for it's designed duration but at some lower throttled pressure. Adding long/slow openings and closings of the valves further agravates the choking.
2 cents worth -------- Bill G.
Looking at what I can get from the data on the GM engine then the cylinder 3.5" dia by 2.6" stroke is 25" cubed. The valve at 1.5" dia is a circumference of 4.7" and with a lift of 0.102" then has an inlet port area of 0.48".
This is then at 15% cutoff a volume of 3.8" roughly of new steam.
So 0.48/3.8 = .12 as a valve area to charge volume factor.
Mine has a valve port area of 0.6 to a charge volume at 15% cutoff of 2.4" cubed.
This yeilds a factor of .25 as a valve area to charge volume factor.
It's a lousy comparison method as other factors should enter in, but it is something. Apples to apples it would be twice the valve area.
The thing to notice is that the Gm engine was/is considered a more modern one yet still took over 20 degrees to finish closing the valve. One which didn't open that far to begin with. With boundary layers of up to 20 thou depending on steam velocitys an opening of 0.102 starts to choke quick. The Williams valves of around 0.050 max lift are almost choked when full open.
These low lifts are expensive when considering that the constant pressure process is not happening at inlet pressure for it's designed duration but at some lower throttled pressure. Adding long/slow openings and closings of the valves further agravates the choking.
2 cents worth -------- Bill G.
|
Re: Williams ws. Rankin October 06, 2006 03:58AM |
Admin Registered: 18 years ago Posts: 2,107 |
HI Bill:
I don't think GM R & D considered the SE-101 to be a particularly advanced steam car, more in the realm of a test bed to look at steam technologies. The boiler is actually much better than I thought based on chatter in the steam community, but it looks like it was drastically under fired. I'd suspect that if pushed it could have likely shown an easy 50% greater output. The transmission was a continuously variable toric design with a truly wild top to bottom range. On the other hand it DID put out 160 hp at 3200 rpm, which is pretty impressive for a first try steam engine, crossheads and all. Conceptually pretty close to the Williamses engines. Everything was way overweight, hard to tell how much would have come out in later models if pursued. As for the valves, the report makes it pretty clear that the cam profiles were chosen so as to ensure that valve train forces were no higher than on then-current production IC engines. This explains the low lift and gentle valve closing. On the plus side, this valve train would likely have held up well in real world service.
Regards,
Ken
I don't think GM R & D considered the SE-101 to be a particularly advanced steam car, more in the realm of a test bed to look at steam technologies. The boiler is actually much better than I thought based on chatter in the steam community, but it looks like it was drastically under fired. I'd suspect that if pushed it could have likely shown an easy 50% greater output. The transmission was a continuously variable toric design with a truly wild top to bottom range. On the other hand it DID put out 160 hp at 3200 rpm, which is pretty impressive for a first try steam engine, crossheads and all. Conceptually pretty close to the Williamses engines. Everything was way overweight, hard to tell how much would have come out in later models if pursued. As for the valves, the report makes it pretty clear that the cam profiles were chosen so as to ensure that valve train forces were no higher than on then-current production IC engines. This explains the low lift and gentle valve closing. On the plus side, this valve train would likely have held up well in real world service.
Regards,
Ken
|
Re: Williams ws. Rankin October 06, 2006 09:53AM |
Registered: 19 years ago Posts: 1,441 |
Bill-Ken,
Since I was in the middle of it, in a minor way, let me say one thing about those GM and Besler cars that they made for that Exhibit of Power GM put on.
They did not want any steam car at all, even then considered them a waste of time because of the lousy fuel milage. They made one and Besler made one for them, just so GM could say "Well we tried and this is the result."
Besler's worked, although more like some squishy Stanley than any hot rod steamer. GM's car had some real design problems and from what Art Underwood told me, was pushed into some storage building and forgotten. As he stated once: "Making two steam cars would show the environmentalists that they tried; but the cars were no good, and to stop their yapping about steam as a viable way to power an automobile."
Steam was the pet idea then, the engine du jour, we have come a long way since then; but to date, no one has shown any advanced and successful steamer that can actually be used as well as any other automobile.
Low lift valves for cutoff are best abandoned, and one should concentrate on constant lift and change the timing, not the lift. One more reason to not consider any high speed steam engine; but one more rational. There is just no logical reason to aim for 3,000-5,000 rpm when no more than 2,000 is very usable. A small increase in displacement to provide the power wanted, is no handicap.
JC
Since I was in the middle of it, in a minor way, let me say one thing about those GM and Besler cars that they made for that Exhibit of Power GM put on.
They did not want any steam car at all, even then considered them a waste of time because of the lousy fuel milage. They made one and Besler made one for them, just so GM could say "Well we tried and this is the result."
Besler's worked, although more like some squishy Stanley than any hot rod steamer. GM's car had some real design problems and from what Art Underwood told me, was pushed into some storage building and forgotten. As he stated once: "Making two steam cars would show the environmentalists that they tried; but the cars were no good, and to stop their yapping about steam as a viable way to power an automobile."
Steam was the pet idea then, the engine du jour, we have come a long way since then; but to date, no one has shown any advanced and successful steamer that can actually be used as well as any other automobile.
Low lift valves for cutoff are best abandoned, and one should concentrate on constant lift and change the timing, not the lift. One more reason to not consider any high speed steam engine; but one more rational. There is just no logical reason to aim for 3,000-5,000 rpm when no more than 2,000 is very usable. A small increase in displacement to provide the power wanted, is no handicap.
JC
|
Re: Williams ws. Rankin October 06, 2006 09:54AM |
Registered: 19 years ago Posts: 835 |
Hi Bill
you might take a look at the piston area in relation to the valve port area.
4' piston x 1' port= 16 to 1 velosity x piston speed. On a long cutoff you can see the preasure drop that occours. the shorter the cutoff the slower the piston speed. Off course shorter cutoff you need higher pressure and higher temp. every thing is a trade off.
Harry
you might take a look at the piston area in relation to the valve port area.
4' piston x 1' port= 16 to 1 velosity x piston speed. On a long cutoff you can see the preasure drop that occours. the shorter the cutoff the slower the piston speed. Off course shorter cutoff you need higher pressure and higher temp. every thing is a trade off.
Harry
|
Re: Williams ws. Rankin October 06, 2006 11:58AM |
Registered: 19 years ago Posts: 1,268 |
Hello Guys,
Truth is, for good analitic comparisons actual measurements, blueprints, cam profiles and inlet steam conditions would have to be known for several well documented engines.
It would probably boil down to what volume of steam is traveling through what size and shape of valve opening under what pressure differential for what time duration and variations of cam profile. Not too very different from IC engines, at least for poppet valves, but just not enough available hard data to make good design judgements from. Steam engine designs also vary quite a bit so that it is often comparing apples to oranges to ponnagramuts.
Still, the biggest valve and port openings with the shortest open/close durations should give the better results.
Best ---- Bill G.
Truth is, for good analitic comparisons actual measurements, blueprints, cam profiles and inlet steam conditions would have to be known for several well documented engines.
It would probably boil down to what volume of steam is traveling through what size and shape of valve opening under what pressure differential for what time duration and variations of cam profile. Not too very different from IC engines, at least for poppet valves, but just not enough available hard data to make good design judgements from. Steam engine designs also vary quite a bit so that it is often comparing apples to oranges to ponnagramuts.
Still, the biggest valve and port openings with the shortest open/close durations should give the better results.
Best ---- Bill G.
|
Re: Williams ws. Rankin October 10, 2006 05:19PM |
Registered: 19 years ago Posts: 1,038 |
Hi Bill
Thinking about your point of shortest open/close durations should be the best. Were you thinking of duration in degrees or time? When I look at open time on my variable cutoff control idea the time duration as about the same through most of the RPM range. It's backwords to old engines where cutoff is shortened it higher speeds. Using cutoff to control the amount of steam means it opens longer for more power.
Andy
Thinking about your point of shortest open/close durations should be the best. Were you thinking of duration in degrees or time? When I look at open time on my variable cutoff control idea the time duration as about the same through most of the RPM range. It's backwords to old engines where cutoff is shortened it higher speeds. Using cutoff to control the amount of steam means it opens longer for more power.
Andy
|
Re: Williams ws. Rankin October 10, 2006 11:55PM |
Registered: 19 years ago Posts: 1,268 |
Hello Andy,
My inlet valve mechanism design is entirely mechanical and coupled to the crankshaft through timeing belts and variable phase mechanisms, modified from a production car, which can change both the advance and closing of the inlet valves. Thus timeing is to degrees of crankshaft rotation. The lift is constant, now at 0.20" effective (more mechanically) and the valves move from close to open or visa-versa in 15 degrees crank. This is at least twice as fast as is normally possible.
For situations of less than 30 degrees from close to open to close again the interactions of the mechanism are the equivalent of changing the "effective" lift. Now at 30 degrees from TDC we are at about 6% cutoff, minus advance less than that. So the lower lifts at these small amounts of cutoff shouldn't be noticed much and in a compound we should be throttling a bit here anyway to stop over expansion.
The design RPM of the inlet valve mechanism is 4K, maybe more.
As I understand it you were working with an electrically actuated valve mechanism. This would then have a more constant timeing from close to open to close again. I seems that the advantage of this would be at the lower RPMs. I researched this a bit and couldn't find any really fast electric valve mechanisms. It wasn't untill I was well into my valve and cam design that I figured out why. The forces are too great for electrics which have to add weight to get more force then build up a stronger field which takes longer to develop, etc. It is a thing chasing it's own tail. Someday maybe it will get there but not now.
Anyway the mechanical system is plenty fast enough for now and, if neccessary, I have a different way altogether to get much higher speeds than 4k. That would though involve it's own development program and would be too much to add to getting the first engine prototypes going. The back burner is all that that idea gets for now.
I have planned a combination of longer cutoff and throttling for lower Rpms and additional sempering through a small valve mechanism to the second stages for really low to starting speeds. This can be run off of the main cams. Remember that even a very inefficient steam engine in a car at low speeds gets amazing economy as long as the rings don't leak. This is why a modern "stuck in a traffic jamb" steam engined car will get some notice in the economy dept.
Any luck with the SES papers? I didn't get much done this weekend. I realized when I was still in my bathrobe at five PM Sunday that I wasn't really feeling too good. Happens more lately it seems, getting old is a bitch. Ended up doing more bearing research on the web. Does anyone have a good formula for the load one roller can safely carry, not a whole complete bearing just one roller between two races?
Thanks and best regards to everyone -------- Bill G.
My inlet valve mechanism design is entirely mechanical and coupled to the crankshaft through timeing belts and variable phase mechanisms, modified from a production car, which can change both the advance and closing of the inlet valves. Thus timeing is to degrees of crankshaft rotation. The lift is constant, now at 0.20" effective (more mechanically) and the valves move from close to open or visa-versa in 15 degrees crank. This is at least twice as fast as is normally possible.
For situations of less than 30 degrees from close to open to close again the interactions of the mechanism are the equivalent of changing the "effective" lift. Now at 30 degrees from TDC we are at about 6% cutoff, minus advance less than that. So the lower lifts at these small amounts of cutoff shouldn't be noticed much and in a compound we should be throttling a bit here anyway to stop over expansion.
The design RPM of the inlet valve mechanism is 4K, maybe more.
As I understand it you were working with an electrically actuated valve mechanism. This would then have a more constant timeing from close to open to close again. I seems that the advantage of this would be at the lower RPMs. I researched this a bit and couldn't find any really fast electric valve mechanisms. It wasn't untill I was well into my valve and cam design that I figured out why. The forces are too great for electrics which have to add weight to get more force then build up a stronger field which takes longer to develop, etc. It is a thing chasing it's own tail. Someday maybe it will get there but not now.
Anyway the mechanical system is plenty fast enough for now and, if neccessary, I have a different way altogether to get much higher speeds than 4k. That would though involve it's own development program and would be too much to add to getting the first engine prototypes going. The back burner is all that that idea gets for now.
I have planned a combination of longer cutoff and throttling for lower Rpms and additional sempering through a small valve mechanism to the second stages for really low to starting speeds. This can be run off of the main cams. Remember that even a very inefficient steam engine in a car at low speeds gets amazing economy as long as the rings don't leak. This is why a modern "stuck in a traffic jamb" steam engined car will get some notice in the economy dept.
Any luck with the SES papers? I didn't get much done this weekend. I realized when I was still in my bathrobe at five PM Sunday that I wasn't really feeling too good. Happens more lately it seems, getting old is a bitch. Ended up doing more bearing research on the web. Does anyone have a good formula for the load one roller can safely carry, not a whole complete bearing just one roller between two races?
Thanks and best regards to everyone -------- Bill G.
|
Re: Williams ws. Rankin October 11, 2006 10:24AM |
Registered: 19 years ago Posts: 1,038 |
Hi Bill
Havn't found the S.E.S. paper as yet. Got busy this weekend. Didn't get the time to go through all my boxes.
In your second paragraph you talked about "throttling a bit here anyway to stop over expansion". Generally throttling is what causes over expansion. Lowering the initial pressure at the start of expansion, lowers the ending pressure. I figure if my valves wind up throttling I would have to extend cutoff to maintain ending pressure. But then I am figuring on expanding to around 5 to 10 PSIA above exhaust or reciever pressures.
The electric valves I am working on, are not powered by electricity. There are two electro-magnets that hold the valve at it's movment extreams. The valve is moved by springs. Unenergized a plate is held centered between two springs under compression. The two electro magnets engage the plate and hold it to one side or the other. The valves have to initially be to one side. The origional design used an initial high curent in one magnet to pull the plate to one side and then swithed to a lower curent. With the plate against the magnet we don't need much curent to hold it there. In operation the holding curent is switched from the magnet holding the plate to the other and the unbalanced spring tension drives the valve toward the other now powered magnet. Initia caries into latching distance and the powered magnet holds it there. Switching the curent to the other magnet and the process gos in the opsite direction to be latched by the other magnet again. The magnet is not the moving force but only holding the valve open or closed.
Andy
Edited 2 time(s). Last edit at 10/11/2006 11:49AM by Andy.
Havn't found the S.E.S. paper as yet. Got busy this weekend. Didn't get the time to go through all my boxes.
In your second paragraph you talked about "throttling a bit here anyway to stop over expansion". Generally throttling is what causes over expansion. Lowering the initial pressure at the start of expansion, lowers the ending pressure. I figure if my valves wind up throttling I would have to extend cutoff to maintain ending pressure. But then I am figuring on expanding to around 5 to 10 PSIA above exhaust or reciever pressures.
The electric valves I am working on, are not powered by electricity. There are two electro-magnets that hold the valve at it's movment extreams. The valve is moved by springs. Unenergized a plate is held centered between two springs under compression. The two electro magnets engage the plate and hold it to one side or the other. The valves have to initially be to one side. The origional design used an initial high curent in one magnet to pull the plate to one side and then swithed to a lower curent. With the plate against the magnet we don't need much curent to hold it there. In operation the holding curent is switched from the magnet holding the plate to the other and the unbalanced spring tension drives the valve toward the other now powered magnet. Initia caries into latching distance and the powered magnet holds it there. Switching the curent to the other magnet and the process gos in the opsite direction to be latched by the other magnet again. The magnet is not the moving force but only holding the valve open or closed.
Andy
Edited 2 time(s). Last edit at 10/11/2006 11:49AM by Andy.
|
Re: Williams ws. Rankin October 11, 2006 10:45PM |
Registered: 19 years ago Posts: 1,268 |
Hello Andy,
True that throttling causes over expansion, but so can a very low cutoff. Remember we are talking here of a high expansion compound. Getting down to the 6% area of cutoff for the first stage out of 30% roughly for full power is starting to get into the low output arena, far less than the 20% power one would think because of the recompression. Meaning that not much steam is passing through the first stage. For a one hundred HP engine the two stages would be putting out about eight HP at design RPM. Now this is where the valves would be just starting to get lower effective lift, so I really would not expect noticable throttling anyway.
It sounds as if your latching valves have a possibility. It occures to me that inductance could slow down the timeing a bit but that by reenergizing latching magnet one as soon after release as possible instead of waiting untill latching magnet two has released the valve, might speed up the process a bit. The magnets could use capacitors to form a tuned circuit for just the right speed to do this. This way when the magnetic latch released it's field would reverse and repel the valve away from it, assuming a strong ceramic magnet for the valve.
Best of fortune with it. ----------- Bill G.
True that throttling causes over expansion, but so can a very low cutoff. Remember we are talking here of a high expansion compound. Getting down to the 6% area of cutoff for the first stage out of 30% roughly for full power is starting to get into the low output arena, far less than the 20% power one would think because of the recompression. Meaning that not much steam is passing through the first stage. For a one hundred HP engine the two stages would be putting out about eight HP at design RPM. Now this is where the valves would be just starting to get lower effective lift, so I really would not expect noticable throttling anyway.
It sounds as if your latching valves have a possibility. It occures to me that inductance could slow down the timeing a bit but that by reenergizing latching magnet one as soon after release as possible instead of waiting untill latching magnet two has released the valve, might speed up the process a bit. The magnets could use capacitors to form a tuned circuit for just the right speed to do this. This way when the magnetic latch released it's field would reverse and repel the valve away from it, assuming a strong ceramic magnet for the valve.
Best of fortune with it. ----------- Bill G.
Sorry, only registered users may post in this forum.
All files from this thread
