Williams ws. Rankin
Posted by Howard Langdon
|
Re: Williams ws. Rankin February 17, 2006 09:09AM |
Registered: 19 years ago Posts: 835 |
|
Re: Williams ws. Rankin February 17, 2006 08:37PM |
Registered: 19 years ago Posts: 318 |
|
Re: Williams ws. Rankin February 18, 2006 04:31AM |
Registered: 19 years ago Posts: 1,453 |
Hi Howard,
Pressure, temp, steam rate, and weight look doable, but my highly experimental idea would be too big (~4'x6', & 2' overhead) for a 350 Chevy engine bay. More advanced experimental stuff is needed. The numbers look good; I bet you can do it. How about something like that in a car (Green Monster II?)? Wow.
Peter
Pressure, temp, steam rate, and weight look doable, but my highly experimental idea would be too big (~4'x6', & 2' overhead) for a 350 Chevy engine bay. More advanced experimental stuff is needed. The numbers look good; I bet you can do it. How about something like that in a car (Green Monster II?)? Wow.
Peter
|
Re: Williams ws. Rankin February 18, 2006 09:13AM |
Registered: 19 years ago Posts: 282 |
Howard,
I couldn't identify your boat class from the info in this post. Is it a ULHAA Rules boat? That has 3 classes with weight limits of 2000, 2500, and 2,700lb.
If there is a limit of 1000 lb for the engine, drive, boiler, water, accessories and fuel there is little to play with - 20 minutes supply of feed water will take it all :-(.
Please contact me direct at gvagg@apex.net.au so some options can be kicked around. You need to get good weight figures for the Williams engine and the gearbox unit, allow for running time fuel and water to see what margin is left for the boiler, feed and fuel system.
Can you afford to run this thing? My first guess for fuel cost is over $US100/hr, and in Australia $US189/hr. Hot rodding has its price.
Graeme
I couldn't identify your boat class from the info in this post. Is it a ULHAA Rules boat? That has 3 classes with weight limits of 2000, 2500, and 2,700lb.
If there is a limit of 1000 lb for the engine, drive, boiler, water, accessories and fuel there is little to play with - 20 minutes supply of feed water will take it all :-(.
Please contact me direct at gvagg@apex.net.au so some options can be kicked around. You need to get good weight figures for the Williams engine and the gearbox unit, allow for running time fuel and water to see what margin is left for the boiler, feed and fuel system.
Can you afford to run this thing? My first guess for fuel cost is over $US100/hr, and in Australia $US189/hr. Hot rodding has its price.
Graeme
|
Re: Williams ws. Rankin February 18, 2006 08:52PM |
Registered: 19 years ago Posts: 318 |
|
Re: Williams ws. Rankin February 20, 2006 07:30PM |
Registered: 19 years ago Posts: 318 |
HIGvegg2.i get the emails back. Wan I send vim. My ad dress is Howard.Langdon@sumpatico.ca
|
Re: Williams ws. Rankin February 24, 2006 07:48PM |
Registered: 19 years ago Posts: 318 |
|
Re: Williams ws. Rankin February 27, 2006 12:39PM |
Registered: 19 years ago Posts: 835 |
Hi Howard,
If you can hold those weights it will be close.should be able to bring down the boiler weight though. The 426 hemi that was run in the 7leter boats weighed in at 750lbs with exaust headers not inc gear box as I ran one in a 7L IPC class. I later ran a 427 chevy which was lighter in a new boat and set a record at the time. Better hull and lighter engine but it would have been faster with the hemi.
The new hemi's are only semi-hemi and a crowned deck where as the old one was a true hemisphere for larger exausts. Loved that engine. sorry for digressing.
Harry
If you can hold those weights it will be close.should be able to bring down the boiler weight though. The 426 hemi that was run in the 7leter boats weighed in at 750lbs with exaust headers not inc gear box as I ran one in a 7L IPC class. I later ran a 427 chevy which was lighter in a new boat and set a record at the time. Better hull and lighter engine but it would have been faster with the hemi.
The new hemi's are only semi-hemi and a crowned deck where as the old one was a true hemisphere for larger exausts. Loved that engine. sorry for digressing.
Harry
|
Re: Williams ws. Rankin February 27, 2006 08:20PM |
Registered: 19 years ago Posts: 318 |
|
Re: Williams ws. Rankin March 05, 2006 07:44PM |
Registered: 19 years ago Posts: 318 |
|
Re: Williams ws. Rankin March 05, 2006 09:25PM |
Registered: 19 years ago Posts: 1,268 |
Hello Howard,
Of course with nothing built and tested giving a water rate would just sound like a lot of bragging or BS. I hope it isn't.
I am shootimg for a water rate of just under 5# for the rankine part of the engine. That is a minimum engine efficiency of around 30% theoretical. Take off for friction and such and it's hard to say. That is at full power so the efficiency should go up as the engine is run at shorter cutoff untill friction takes over.
When the Joule cycle is added with the turbine the efficiency shoots through the roof. The turbine runs at around 2000 degrees F at fairly low pressure and the rejected heat goes to the boiler. At full power the turbine is adding an extra 50%. The design target for the piston engine is 200 HP at 1800 RPM, The turbine will add another 100 HP.
This is only a rough outline even theoretically, as it's been hard to get the dedicated time and energy neccessary to keep up a good design pace. Still it all seems to fall into place, no bugaboos.
Best ------------ Bill G.
Of course with nothing built and tested giving a water rate would just sound like a lot of bragging or BS. I hope it isn't.
I am shootimg for a water rate of just under 5# for the rankine part of the engine. That is a minimum engine efficiency of around 30% theoretical. Take off for friction and such and it's hard to say. That is at full power so the efficiency should go up as the engine is run at shorter cutoff untill friction takes over.
When the Joule cycle is added with the turbine the efficiency shoots through the roof. The turbine runs at around 2000 degrees F at fairly low pressure and the rejected heat goes to the boiler. At full power the turbine is adding an extra 50%. The design target for the piston engine is 200 HP at 1800 RPM, The turbine will add another 100 HP.
This is only a rough outline even theoretically, as it's been hard to get the dedicated time and energy neccessary to keep up a good design pace. Still it all seems to fall into place, no bugaboos.
Best ------------ Bill G.
|
Re: Williams ws. Rankin March 06, 2006 11:51AM |
Registered: 19 years ago Posts: 1,038 |
Hi Howard.
I have a simplistic simulater done in mathconex and mathcad. Due to bugs in the interface between the two programs that causes memory leaks that cause it to soon run out of memory I was only able ot do a few revoloutions. I am working on one in vissim using (IAPWS-95 scientific steam property function I am writing) I am about half way through the neccessary function. I have 9 of the 12 dimensionless Helmholtz free energy and it'ts derivative functions working. There two parts to the helmholtz energy equations. These are designated by the Greek leter phi (Looks like a small o with a vertical line through it) with a supper script r or o. Can't use Greek leters here so call them phir and phio.
Phio is for the ideal gass part and phir is the real gas part. These function are in terms of delta and tau More Greek leters. delta is basicly density and tau is temperature. But to be real percise they are dimensionless quanties notmalized to the critical point.
tau = (critical point temperature)/(absolut temperature)
delta = density/(critical point density)
phio = ln(delta) + no1 + no2 * tau + n03* ln(tau) + sum{i=4..8}(no * ln(1-e^(-lambda*tau)))
phir=sum{i=1..7}(nr*delta^d) + sum{i=8..51}(n*delta^d*tau^t*e^-(delta^c)) + sum{i=52..54}(n*delta^d*tau^t*e^(-alpha*(delta-epsilon)^2 - Beta*(tau-lambda)^2)) + sum{i=55..56}(nDelta^b*delta*psi)
Delta = theta^2 + B*((delta-1)^2)^a
theta = (1-tau) + A*((delta-1)^2)^(1/(2*Beta))
psi = e^(-C(delta-1)^2 - D*(tau-1)^2)
The twelve formula are:
The above 2 equation above. (2 formula phio(delta, tau) and phir(delta, tau))
Their first partail derivitive with respect delta. (2 formula dphiodelta(delta, tau) and dphiRdelta(delta, tau))
Their second partail derivitive with respect delta. (2 formula dphiodeltadelta(delta, tau) and dphirdeltadelta(delta, tau))
Their first partail derivitive with respect tau. (2 formula dphiotau(delta, tau) and dphirtau(delta, tau))
Their second partail derivitive with respect tau. (2 formula dphiotautau(delta, tau) and dphirtautau(delta, tau))
Their partial derivites with respect to delta and tau. (2 formula dphiodeltatau(delta, tau) and dphirdeltatau(delta, tau))
phio(delta, tau) and phir(delta, tau))
dphiodelta(delta, tau) and dphiRdelta(delta, tau))
dphiodeltadelta(delta, tau) and dphirdeltadelta(delta, tau))
dphiotau(delta, tau) and dphirtau(delta, tau))
dphiotautau(delta, tau) and dphirtautau(delta, tau))
dphiodeltatau(delta, tau) and dphirdeltatau(delta, tau))
I am working on dphirdeltadelta(delta, tau)). Have dphirtautau(delta, tau)) and dphirdeltatau(delta, tau)) left t do.
The other steam properties are then calculated from these:
rho is mass density kg/m^3 R = 0.46151805 T is temperatue degrees K dphirdelta is dphirdelta(delta, tau) other phir and phio function will be done the same for shortness of expersions.
presure = (1 + delta*dphirdelta) * rho * R * T
internal energy = tau*(dphiotau + dphirtau) * R * T
enthalpy = (1 + tau*((dphiotau + dphirtau)) + delta*dphirdelta) * R * T
entropy = (tau*(dphiotau + dphirtau) + phio + phir) * R
Isochoric heat capicity Cv = -tau^2*(dphiotautau + dphirtautau) * R
Isobaric heat capicity Cp = -tau^2*(dphiotautau + dphirtautau) + (1 + delta*dphirdelta-delta*tau*dphirdeltatau)^2/(1+2*delta*dphirdelta + delta^2*phirdeltadelta)
The six or more properties (speed of sound), Isothermal throttling coefficient, Isentropic temperature pressure coefficient etc I am not implementing at thi time.
I have pressure, internal energy, enthapy and entropy implemented as well.
Have a funcion that solves for specific volume given temperature and pressure. Working on other solvers for pressure and entropy, volume and entropy etc. I am working on the second order derivative functions so that I can calculate Cp and Cv. Hope to use them in the expansion solvers. Knowing Cp Cv and k will enable me to make a close first guesses when doing small expansions. Visim is a simulater that calculates aproxamations for small time steps.
Still problems using mathcad and vissim together. I tried vissim with mathcad doing the steam calculations and ran into problems with it hanging. This will be much much much... better. The speed of native Vissim blocks is a big improvement over the slow program to program mesage system using mathcad. The IAPWS-95 formulations are the most accurate formulations. IFC-97 is faster but is less accurate and it divides up the topology into 5 regions with different property formula that have inconsistant boundry line values. That is that when you cross a broundry the calculations are done with different formula and the value is not a smothe transition. The IFC-67 is the same way having 4 regions. The 5 region in the IFC-97 is an extended pressure region that the IFC-67 didn't have.
My simulater using mathconex was simplistic in that it, like an ideal cycle, did not include losses. Mixing heat transfers, of steam masses, were calculated. So some losses are done.
To get to your question the simultor didn't produce steam rates. though it could calculate it. MPG was calculated. So the bottom line is around 48 MPG at 60 MPH. I couldn't run very high initial first stage temperature with out running into the IFC-67 upper temperature limit on compression. I was limited to around 800F. I figure that with IAPWS-95, which has extreamly high upper pressure and temperature limits, compression calculations will be no problem.
My design is a three stage counterflow high compression compound. The clearance and displacement of all are stagesis variable. The displacemen and clearance are tied together in that clearance is varied by changing the effective stroke of the engine. All stages are tied together. The exact relation of clearance to displacement isdependent in it initial/or min clearance. Valve timming of exhaust and inlet and for each stage are independent. Basicly for variable power output the clearance is varied, The displacement also varies with clearance. The inlet cutoff is set to expand to just above (around 15 PSI) of interstage reciever pressure set point or around 5 PSI above exhaust pressure on the last stage. When the interstage reciever pressure is below the set point make up steam from the engine inlet is regulated into the reciever. It min pressure is it set point. The engine is designed such that across the entire power bound each succesive stage uses a little more steam the the previous. The steam regulated into the reciever in low pressure is make up steam. The engiine is designed so that each stage produces about equal torque across the power band.
The combination of the high compression and the supperheating effect of the makeup steam in the reciever maintains good quality steam through all stages. I am shooting for 90% quality exhaust. Slightly wet exhaust. The simulater achieved 88% quality at 750F initial inlet steam.
I use a mathcad work sheet that calculates 1000s of cycles across a clearance range that comes up with an initial valve timming table. That goes into the sim where I can fine tune the valve timmings further. Hopfully I can get a test rig LAB engine built to evauate my simulater against and further refine the simulater. Scalling is the next objective. Getting the sim to match up with diferent engine sizes.
The goal is to have a simulater that closely matches the real world and can be used as a design tool. I hope to eventually be able to eneter engine parameter of hourse power at some RPM ... etc and have all the pertant engine parameters calculated.
The simulater is made up of blocks. User defined blocks can be created from blocks. These can be leveled like in programming. Vissim is basicly a special graphical programming linguage for writing simulaters.
My simulater will be made up of some high level blocks like expander blocks, reviever blocks and etc. An expander block takes parameter inputs of bore, stroke, min clearance, clearance-displacement control, inlet and exhaust valve timming. It inputs an inlet steam point and outputs inlet steam flow mass, exhaust steam flow mass and steam properties.
The high level blocks can be conected into different engine configurations.
It's a big project timewise.
Andy
Andy
I have a simplistic simulater done in mathconex and mathcad. Due to bugs in the interface between the two programs that causes memory leaks that cause it to soon run out of memory I was only able ot do a few revoloutions. I am working on one in vissim using (IAPWS-95 scientific steam property function I am writing) I am about half way through the neccessary function. I have 9 of the 12 dimensionless Helmholtz free energy and it'ts derivative functions working. There two parts to the helmholtz energy equations. These are designated by the Greek leter phi (Looks like a small o with a vertical line through it) with a supper script r or o. Can't use Greek leters here so call them phir and phio.
Phio is for the ideal gass part and phir is the real gas part. These function are in terms of delta and tau More Greek leters. delta is basicly density and tau is temperature. But to be real percise they are dimensionless quanties notmalized to the critical point.
tau = (critical point temperature)/(absolut temperature)
delta = density/(critical point density)
phio = ln(delta) + no1 + no2 * tau + n03* ln(tau) + sum{i=4..8}(no * ln(1-e^(-lambda*tau)))
phir=sum{i=1..7}(nr*delta^d) + sum{i=8..51}(n*delta^d*tau^t*e^-(delta^c)) + sum{i=52..54}(n*delta^d*tau^t*e^(-alpha*(delta-epsilon)^2 - Beta*(tau-lambda)^2)) + sum{i=55..56}(nDelta^b*delta*psi)
Delta = theta^2 + B*((delta-1)^2)^a
theta = (1-tau) + A*((delta-1)^2)^(1/(2*Beta))
psi = e^(-C(delta-1)^2 - D*(tau-1)^2)
The twelve formula are:
The above 2 equation above. (2 formula phio(delta, tau) and phir(delta, tau))
Their first partail derivitive with respect delta. (2 formula dphiodelta(delta, tau) and dphiRdelta(delta, tau))
Their second partail derivitive with respect delta. (2 formula dphiodeltadelta(delta, tau) and dphirdeltadelta(delta, tau))
Their first partail derivitive with respect tau. (2 formula dphiotau(delta, tau) and dphirtau(delta, tau))
Their second partail derivitive with respect tau. (2 formula dphiotautau(delta, tau) and dphirtautau(delta, tau))
Their partial derivites with respect to delta and tau. (2 formula dphiodeltatau(delta, tau) and dphirdeltatau(delta, tau))
phio(delta, tau) and phir(delta, tau))
dphiodelta(delta, tau) and dphiRdelta(delta, tau))
dphiodeltadelta(delta, tau) and dphirdeltadelta(delta, tau))
dphiotau(delta, tau) and dphirtau(delta, tau))
dphiotautau(delta, tau) and dphirtautau(delta, tau))
dphiodeltatau(delta, tau) and dphirdeltatau(delta, tau))
I am working on dphirdeltadelta(delta, tau)). Have dphirtautau(delta, tau)) and dphirdeltatau(delta, tau)) left t do.
The other steam properties are then calculated from these:
rho is mass density kg/m^3 R = 0.46151805 T is temperatue degrees K dphirdelta is dphirdelta(delta, tau) other phir and phio function will be done the same for shortness of expersions.
presure = (1 + delta*dphirdelta) * rho * R * T
internal energy = tau*(dphiotau + dphirtau) * R * T
enthalpy = (1 + tau*((dphiotau + dphirtau)) + delta*dphirdelta) * R * T
entropy = (tau*(dphiotau + dphirtau) + phio + phir) * R
Isochoric heat capicity Cv = -tau^2*(dphiotautau + dphirtautau) * R
Isobaric heat capicity Cp = -tau^2*(dphiotautau + dphirtautau) + (1 + delta*dphirdelta-delta*tau*dphirdeltatau)^2/(1+2*delta*dphirdelta + delta^2*phirdeltadelta)
The six or more properties (speed of sound), Isothermal throttling coefficient, Isentropic temperature pressure coefficient etc I am not implementing at thi time.
I have pressure, internal energy, enthapy and entropy implemented as well.
Have a funcion that solves for specific volume given temperature and pressure. Working on other solvers for pressure and entropy, volume and entropy etc. I am working on the second order derivative functions so that I can calculate Cp and Cv. Hope to use them in the expansion solvers. Knowing Cp Cv and k will enable me to make a close first guesses when doing small expansions. Visim is a simulater that calculates aproxamations for small time steps.
Still problems using mathcad and vissim together. I tried vissim with mathcad doing the steam calculations and ran into problems with it hanging. This will be much much much... better. The speed of native Vissim blocks is a big improvement over the slow program to program mesage system using mathcad. The IAPWS-95 formulations are the most accurate formulations. IFC-97 is faster but is less accurate and it divides up the topology into 5 regions with different property formula that have inconsistant boundry line values. That is that when you cross a broundry the calculations are done with different formula and the value is not a smothe transition. The IFC-67 is the same way having 4 regions. The 5 region in the IFC-97 is an extended pressure region that the IFC-67 didn't have.
My simulater using mathconex was simplistic in that it, like an ideal cycle, did not include losses. Mixing heat transfers, of steam masses, were calculated. So some losses are done.
To get to your question the simultor didn't produce steam rates. though it could calculate it. MPG was calculated. So the bottom line is around 48 MPG at 60 MPH. I couldn't run very high initial first stage temperature with out running into the IFC-67 upper temperature limit on compression. I was limited to around 800F. I figure that with IAPWS-95, which has extreamly high upper pressure and temperature limits, compression calculations will be no problem.
My design is a three stage counterflow high compression compound. The clearance and displacement of all are stagesis variable. The displacemen and clearance are tied together in that clearance is varied by changing the effective stroke of the engine. All stages are tied together. The exact relation of clearance to displacement isdependent in it initial/or min clearance. Valve timming of exhaust and inlet and for each stage are independent. Basicly for variable power output the clearance is varied, The displacement also varies with clearance. The inlet cutoff is set to expand to just above (around 15 PSI) of interstage reciever pressure set point or around 5 PSI above exhaust pressure on the last stage. When the interstage reciever pressure is below the set point make up steam from the engine inlet is regulated into the reciever. It min pressure is it set point. The engine is designed such that across the entire power bound each succesive stage uses a little more steam the the previous. The steam regulated into the reciever in low pressure is make up steam. The engiine is designed so that each stage produces about equal torque across the power band.
The combination of the high compression and the supperheating effect of the makeup steam in the reciever maintains good quality steam through all stages. I am shooting for 90% quality exhaust. Slightly wet exhaust. The simulater achieved 88% quality at 750F initial inlet steam.
I use a mathcad work sheet that calculates 1000s of cycles across a clearance range that comes up with an initial valve timming table. That goes into the sim where I can fine tune the valve timmings further. Hopfully I can get a test rig LAB engine built to evauate my simulater against and further refine the simulater. Scalling is the next objective. Getting the sim to match up with diferent engine sizes.
The goal is to have a simulater that closely matches the real world and can be used as a design tool. I hope to eventually be able to eneter engine parameter of hourse power at some RPM ... etc and have all the pertant engine parameters calculated.
The simulater is made up of blocks. User defined blocks can be created from blocks. These can be leveled like in programming. Vissim is basicly a special graphical programming linguage for writing simulaters.
My simulater will be made up of some high level blocks like expander blocks, reviever blocks and etc. An expander block takes parameter inputs of bore, stroke, min clearance, clearance-displacement control, inlet and exhaust valve timming. It inputs an inlet steam point and outputs inlet steam flow mass, exhaust steam flow mass and steam properties.
The high level blocks can be conected into different engine configurations.
It's a big project timewise.
Andy
Andy
|
Re: Williams ws. Rankin March 06, 2006 12:37PM |
Registered: 19 years ago Posts: 835 |
|
Re: Williams ws. Rankin March 07, 2006 07:47PM |
Registered: 19 years ago Posts: 318 |
|
Re: Williams ws. Rankin March 08, 2006 01:34AM |
Registered: 19 years ago Posts: 1,268 |
Hello Howard and everyone.
I need to correct a number in my above post, sorry. At full tilt if the engine is getting 30% efficiency it is at a 5.9# water rate not 5#. The 5# rate would happen at a 35% engine efficiency, shorter cutoff and just below about half power. The overall expansion ratio for the compound at this range would be 40 to 50. At lower outputs and greater expansion ratios the efficiency should still go up a little, too early to tell what friction and such will do.
Howard with an input of 1000 psia at 1050 F the Williams would have to get an engine efficiency of 37% for a 5# water rate and would have to expand down into wet steam about to about 90%. The difference is because water rate and efficiency are not related linearly. The more superheat the less efficient an engine has to be to get the same water rate as a lower temperature engine.
An exhaust heat exchanger will help the overall efficiency somewhat but not near as much as expanding the steam through the engine would.
The problem as I see it is still that a single stage engine just can't get the high expansion ratios neccessary for high efficiencys and put out much power at short cutoffs. I would favor a Williams in medium to long cutoff exhausting into a turbine.
Of course at the same expansion ratios the single stage engine would put out less HP than a compound but the efficiency should be better because of the smaller size and everything.
Anyway best of luck ---------- Bill G.
Edited 1 time(s). Last edit at 03/08/2006 01:40AM by Bill Gatlin.
I need to correct a number in my above post, sorry. At full tilt if the engine is getting 30% efficiency it is at a 5.9# water rate not 5#. The 5# rate would happen at a 35% engine efficiency, shorter cutoff and just below about half power. The overall expansion ratio for the compound at this range would be 40 to 50. At lower outputs and greater expansion ratios the efficiency should still go up a little, too early to tell what friction and such will do.
Howard with an input of 1000 psia at 1050 F the Williams would have to get an engine efficiency of 37% for a 5# water rate and would have to expand down into wet steam about to about 90%. The difference is because water rate and efficiency are not related linearly. The more superheat the less efficient an engine has to be to get the same water rate as a lower temperature engine.
An exhaust heat exchanger will help the overall efficiency somewhat but not near as much as expanding the steam through the engine would.
The problem as I see it is still that a single stage engine just can't get the high expansion ratios neccessary for high efficiencys and put out much power at short cutoffs. I would favor a Williams in medium to long cutoff exhausting into a turbine.
Of course at the same expansion ratios the single stage engine would put out less HP than a compound but the efficiency should be better because of the smaller size and everything.
Anyway best of luck ---------- Bill G.
Edited 1 time(s). Last edit at 03/08/2006 01:40AM by Bill Gatlin.
|
Re: Williams ws. Rankin March 08, 2006 09:18AM |
Registered: 19 years ago Posts: 835 |
Hi Bill
What you are refering to is the net water rate not after losses to drive pumps boiler eff ect. I belive in the real world you will find that the losses will eat up the compound engine efficency.The higher super heat needed to make up the losses thru the valves and the radiant area and a more complacated machine. I studed this for along time and designed an engine with compound exp. in a single cyl. by admission on the bottom side with a large diameter core fore the highpressure stage and a bumpvalve to the top side for the low pressure. Simple but in the end no real gain than to have a small bore to a long stroke to allow more expansions with more superheat High super heat can lower boiler eff. why we went to super critical as the boiler is the first heat exchanger and in a modern engine 90% in this area is nessary. Spliting hairs on the efficency might hurt if the cost to manufacture the machine is to high. A large diesel eng manufactuer told us if the eng cost 20% more it will not be used.
Harry
What you are refering to is the net water rate not after losses to drive pumps boiler eff ect. I belive in the real world you will find that the losses will eat up the compound engine efficency.The higher super heat needed to make up the losses thru the valves and the radiant area and a more complacated machine. I studed this for along time and designed an engine with compound exp. in a single cyl. by admission on the bottom side with a large diameter core fore the highpressure stage and a bumpvalve to the top side for the low pressure. Simple but in the end no real gain than to have a small bore to a long stroke to allow more expansions with more superheat High super heat can lower boiler eff. why we went to super critical as the boiler is the first heat exchanger and in a modern engine 90% in this area is nessary. Spliting hairs on the efficency might hurt if the cost to manufacture the machine is to high. A large diesel eng manufactuer told us if the eng cost 20% more it will not be used.
Harry
|
Re: Williams ws. Rankin March 08, 2006 11:55AM |
Registered: 19 years ago Posts: 1,268 |
Hello Harry,
Correct, net water rate. The real world losses vary too much with load to work out untill the design is closer to finalization so I'm just allowing 5% for friction and a 5% loss in work produced during expansion which ends up as extra heat at the end of expansion. Same in reverse for recompression.
Again though, if a compound doesn't have such a small transfer port volume between stages to allow full second stage recompression as mine does, or another compensating mechanism such as Andys does with additional steam between stages, there is little hope of bettering single stage efficiency.
Without second stage recompression as well as first stage the compounds efficiency falls apart at anything much less than full power.
As far as costs and scope of project imagine designing and building a modern diesel engine from scratch, without much usable prior art to go on other than what was being done in 1910. Of course I guess you of all people don't have to imagine that too hard. LOL The costs or producing a finished product will go down as the engine advances from prototypes to an on the road product. Right now it's just get something to put on the test bench, prove the thing out and make it work reliably.
A modern diesel engine that gets the perported 40% efficiency starts with 56% theoretical thermodynamic efficiency, thats 71% of its theoretical. That fits in somewhere.
I am hoping that from the design there are a lot of offshoots that may be valuable in themselves. I think the cam system will work well for most steam engines, it should be about the fastest reliable mechanical poppet valve system, the crank box will be great for compressors especially those great big ones with cross heads and possibly diesel engines, and the Joule cycle side is a totally different engine on its own.
By the way, the Joule cycle shares the second stage double acting pistons so mechanical losses from having a second stage are minimised.
Anyway ain't gonna quit.
Thanks for the tip about Ceradine, It amazes me how some companies hate little people developing uses for their products. Won't mortgage the house for them.
Wishing you the best of days ------------- Bill G.
Correct, net water rate. The real world losses vary too much with load to work out untill the design is closer to finalization so I'm just allowing 5% for friction and a 5% loss in work produced during expansion which ends up as extra heat at the end of expansion. Same in reverse for recompression.
Again though, if a compound doesn't have such a small transfer port volume between stages to allow full second stage recompression as mine does, or another compensating mechanism such as Andys does with additional steam between stages, there is little hope of bettering single stage efficiency.
Without second stage recompression as well as first stage the compounds efficiency falls apart at anything much less than full power.
As far as costs and scope of project imagine designing and building a modern diesel engine from scratch, without much usable prior art to go on other than what was being done in 1910. Of course I guess you of all people don't have to imagine that too hard. LOL The costs or producing a finished product will go down as the engine advances from prototypes to an on the road product. Right now it's just get something to put on the test bench, prove the thing out and make it work reliably.
A modern diesel engine that gets the perported 40% efficiency starts with 56% theoretical thermodynamic efficiency, thats 71% of its theoretical. That fits in somewhere.
I am hoping that from the design there are a lot of offshoots that may be valuable in themselves. I think the cam system will work well for most steam engines, it should be about the fastest reliable mechanical poppet valve system, the crank box will be great for compressors especially those great big ones with cross heads and possibly diesel engines, and the Joule cycle side is a totally different engine on its own.
By the way, the Joule cycle shares the second stage double acting pistons so mechanical losses from having a second stage are minimised.
Anyway ain't gonna quit.
Thanks for the tip about Ceradine, It amazes me how some companies hate little people developing uses for their products. Won't mortgage the house for them.
Wishing you the best of days ------------- Bill G.
|
Re: Williams ws. Rankin March 08, 2006 01:35PM |
Registered: 19 years ago Posts: 835 |
Bill
You are right, follow your dream,,didnot mean to rain on your parade. especialy sence I have not seen what you have. I was bad mouthed last year by a self proclamed expert who new nouthing about a modern engine. of course I would like a spirited debate. He said it would not run, du. LOL So much for the expert.
happy steaming
Harry
You are right, follow your dream,,didnot mean to rain on your parade. especialy sence I have not seen what you have. I was bad mouthed last year by a self proclamed expert who new nouthing about a modern engine. of course I would like a spirited debate. He said it would not run, du. LOL So much for the expert.
happy steaming
Harry
|
Re: Williams ws. Rankin March 08, 2006 02:08PM |
Registered: 19 years ago Posts: 1,038 |
Hi Herry.
A compound of equal output should be a lot smaller, have reasionable valve timming and clearances.
The problem with a single stage operating at close to full expamsion is simply the verry short valve timming and clearance your engine would have to have.
As an example starting with 1480 PSIA inlet steam at 950F.
Expand it down to 16 PSIA. Slightly above 1 Atm(14.696 PSIA).
The expansion ratio is 40.8:1 In order to get that in a single stage I used a clearance volume of only 2%. Probably not atainable. The cutoff point is 7.3 degrees after TDC. 2.028% of a revoloution. And 0.50% of the stroke. Now how much time is the inlet open at 1000 RPM. 2.028% of 0.060 seconds. You lnlit valve must open and close in about 0.001217 seconds.
A compound is going to scrifice some expansion to interstage pressure drops. My three stage compound operating between 1480 PSIA and 15 PSIA has a combined expansion around 28:1. There is non-work pressure drop expansions into the recievers and to the exhaust.
So an equiivant expansion in a single stage of 28:1 would be more equiivant.
A cycle of 1000 PSIA 950 F expansing to 16PSIA comes close. It has an expansion ratio of 27.84:1. In order to achieve that I used a clearance of 3% that would be hard to achieve. That allowed a 8.6 degree cutoff. 0.7% of stroke. That is a little better time wise. At 1000 RPM the valve has 0.001433 seconds to get open and close.
Note the cutoff and clearance values of the single expansion engine. With high compression. About 21.6% of the steam in the cycle is inlet steamm. With the compound having 5% clearance 3.03 expansion ration 85.7% of the steam is inlet steam. 5% cutoff and 3.03:1 exoansion is a cutoff of 59.8 degrees 29.6% of stroke.
Assumming equal output for equal steam quanty in. A single expansion engine will need a displacement of 1.5 times that of the compound. That differance comes from the clearance space to cutoff ratio. The relative inlet steam amounts of each engine. 21.6% in the single stage expansion to 85.7% in the 3 stage expansion. We need the single stage piston to be 29.3/0.7=41.857143 times the compounds first stage. With an expansion ratio of 3.03 for each stage The compounds displacement of 3.03^3 times the first stage displacement. So the engine size ratio is (29.3/0.7)/3.03^3=1.504671
The single stage expansion 3% clearance 8.6 degree cutoff and valve open time of 0.001433 seconds against 3 stage compound having 5% clearance 59.8 degrees cutoff and a valve open time of 0.009967 seconds and 2/3 the total displacement.
Andy
A compound of equal output should be a lot smaller, have reasionable valve timming and clearances.
The problem with a single stage operating at close to full expamsion is simply the verry short valve timming and clearance your engine would have to have.
As an example starting with 1480 PSIA inlet steam at 950F.
Expand it down to 16 PSIA. Slightly above 1 Atm(14.696 PSIA).
The expansion ratio is 40.8:1 In order to get that in a single stage I used a clearance volume of only 2%. Probably not atainable. The cutoff point is 7.3 degrees after TDC. 2.028% of a revoloution. And 0.50% of the stroke. Now how much time is the inlet open at 1000 RPM. 2.028% of 0.060 seconds. You lnlit valve must open and close in about 0.001217 seconds.
A compound is going to scrifice some expansion to interstage pressure drops. My three stage compound operating between 1480 PSIA and 15 PSIA has a combined expansion around 28:1. There is non-work pressure drop expansions into the recievers and to the exhaust.
So an equiivant expansion in a single stage of 28:1 would be more equiivant.
A cycle of 1000 PSIA 950 F expansing to 16PSIA comes close. It has an expansion ratio of 27.84:1. In order to achieve that I used a clearance of 3% that would be hard to achieve. That allowed a 8.6 degree cutoff. 0.7% of stroke. That is a little better time wise. At 1000 RPM the valve has 0.001433 seconds to get open and close.
Note the cutoff and clearance values of the single expansion engine. With high compression. About 21.6% of the steam in the cycle is inlet steamm. With the compound having 5% clearance 3.03 expansion ration 85.7% of the steam is inlet steam. 5% cutoff and 3.03:1 exoansion is a cutoff of 59.8 degrees 29.6% of stroke.
Assumming equal output for equal steam quanty in. A single expansion engine will need a displacement of 1.5 times that of the compound. That differance comes from the clearance space to cutoff ratio. The relative inlet steam amounts of each engine. 21.6% in the single stage expansion to 85.7% in the 3 stage expansion. We need the single stage piston to be 29.3/0.7=41.857143 times the compounds first stage. With an expansion ratio of 3.03 for each stage The compounds displacement of 3.03^3 times the first stage displacement. So the engine size ratio is (29.3/0.7)/3.03^3=1.504671
The single stage expansion 3% clearance 8.6 degree cutoff and valve open time of 0.001433 seconds against 3 stage compound having 5% clearance 59.8 degrees cutoff and a valve open time of 0.009967 seconds and 2/3 the total displacement.
Andy
|
Re: Williams ws. Rankin March 08, 2006 02:32PM |
Registered: 19 years ago Posts: 835 |
Hi Andy
Our 6cyl eng is designed to operate at 3600rpm the valve train is mostly titanium and the timing duration and cutoff and clearance volume are all varible and are rpm dependent, however can be operated indepedently. the 1 cyc bump valve eng was run at 7000rpm, every one scatered.but it did not break at least not that time. amazing how smooth it became. We will bring the 6 cyl to the next SACA meet.
Harry
Our 6cyl eng is designed to operate at 3600rpm the valve train is mostly titanium and the timing duration and cutoff and clearance volume are all varible and are rpm dependent, however can be operated indepedently. the 1 cyc bump valve eng was run at 7000rpm, every one scatered.but it did not break at least not that time. amazing how smooth it became. We will bring the 6 cyl to the next SACA meet.
Harry
|
Re: Williams ws. Rankin March 08, 2006 02:35PM |
Registered: 19 years ago Posts: 1,268 |
Thanks Harry,
There is so much more to a developing engine concept than should be shared in an open forum. Not if there is much chance of commercial intrest.
I am sure, for instance, that I have a general concept of how your engine works but the finer details of it's workings which are not patent applied for or are in development probably are not open to the public, or weren't in the early stages. Trying to explain how something works while leaving something out sometimes comes across as silly or unreasonable. I am sure my ideas have at times sounded that way. I can promise they will in the future also.
My great hope is to get my stuff to the test bench phase. So much more can be done with real world results in hand than just paper. And by the way, even though I have studied the subject a lot and put a good deal of energy into it I don't consider myself an expert. George Nutz, Jim Crank, Jerry Peoples, and a very few others, these guys are the experts.
Be Well ----------- Bill G.
There is so much more to a developing engine concept than should be shared in an open forum. Not if there is much chance of commercial intrest.
I am sure, for instance, that I have a general concept of how your engine works but the finer details of it's workings which are not patent applied for or are in development probably are not open to the public, or weren't in the early stages. Trying to explain how something works while leaving something out sometimes comes across as silly or unreasonable. I am sure my ideas have at times sounded that way. I can promise they will in the future also.
My great hope is to get my stuff to the test bench phase. So much more can be done with real world results in hand than just paper. And by the way, even though I have studied the subject a lot and put a good deal of energy into it I don't consider myself an expert. George Nutz, Jim Crank, Jerry Peoples, and a very few others, these guys are the experts.
Be Well ----------- Bill G.
|
Re: Williams ws. Rankin March 08, 2006 03:05PM |
Registered: 19 years ago Posts: 1,038 |
Herry
What is your cutoff, inlet pressure and clearance at 3600 RPM? In my last post I explained the problem. The clearance and cutoff from even 1000 PSIA to get full expansion is just impossable to do. 3% clearance and 0.7% cutoff. Less clearance more cutoff but you are still looking at values too small for either. How do you explain this. I trust your engine runs. So how are you getting around these problems? Or are you?
Andy
What is your cutoff, inlet pressure and clearance at 3600 RPM? In my last post I explained the problem. The clearance and cutoff from even 1000 PSIA to get full expansion is just impossable to do. 3% clearance and 0.7% cutoff. Less clearance more cutoff but you are still looking at values too small for either. How do you explain this. I trust your engine runs. So how are you getting around these problems? Or are you?
Andy
|
Re: Williams ws. Rankin March 09, 2006 09:07AM |
Registered: 19 years ago Posts: 835 |
|
Re: Williams ws. Rankin March 09, 2006 12:15PM |
Registered: 19 years ago Posts: 1,771 |
Harry I use some secret water also. See attachments.
|
Re: Williams ws. Rankin March 09, 2006 02:20PM |
Registered: 19 years ago Posts: 835 |
|
Re: Williams ws. Rankin March 09, 2006 03:02PM |
Registered: 19 years ago Posts: 375 |
|
Re: Williams ws. Rankin March 09, 2006 04:51PM |
Registered: 19 years ago Posts: 1,038 |
Hi Herry
I thought you might see an error in my equal power calculation. I came up with a single expansion engine needing to be about 150% larger then a compound to produce the same power. You might look that part over. I figur the amount of inlet steam must be about equal to produce the same power. The single stage is clearance limited. I use 3% but beleave that to be unobtainable. That made the cutoff to be 0.7% to get expansion to 16 PSIA from 1000 PSIA. The three stage compound with 5% clearance and 29.6% cutoff, 3 low expansion of 3.03, to expand to 16 PSIA from 1480 PSIA. I then figured the relative volume of the single sage to the compound first stage so they would have equal quanties of inlet steam. Figured the rest of the compound stages and came up with the 1.5 number. It seamed wrong. The overall isentropic expansion is very close on both engine though the compound operates at a higher first stage inlet pressure. That higher pressure allows for isenthalpic pressure drops between stages that do no work. The over all thoritical cycle efficiency is about the same. So equal amounts of steam would produce equal amounts of power.
The thing is you would not be able run a single stage engine with the small clearances and cutoff. At least I think it would be impossable. The small clearance is required to get the compression and expansion ratio. With a high compression engine, the smaller the clearance the more power you can make for a given displacement. The smaller the clearance the longer the cutoff for the same expansion. Longer cutoff, more steam going through the engine. more power.
I used percentages of steam from my steam cycle calculator. But doing it by cutoffs gets close. We have 0.7% cutoff for single expansion to 29.6% cutoff on firstage of compound. To simplify assume all stage to be the same 29.6% Or haveing a 3:1 expansion ratio. The total volume of the compound would volume v1 of first stage plus volume v2 of second stage and v3 volume of thord stage.
Vc = V1+v2+v3 wher v2 = 3 * v1 and v3 = 3*v2 = 9*v1
vc = v1 + 3*v1 + 9*v1 = 13*v1
The volume of the single expansion Vs is figured so that 0.7% of Vs is equal to 29.6 % of v1
0.07 * Vs = 0.296 * V1
Vs = 0.296/0.07 * V1 = 42.286 V1
Vs/Vc = (42.286 V1) / (13*v1) = 3.253
Vs = 3.253 * Vc
Using steam densities and calculated percentages of inlet steam going the rough a compression cycle I calculated it to be a 1.5:1.
The end result. It seams counter intutitive that the compound could be snaller.
Andy
I thought you might see an error in my equal power calculation. I came up with a single expansion engine needing to be about 150% larger then a compound to produce the same power. You might look that part over. I figur the amount of inlet steam must be about equal to produce the same power. The single stage is clearance limited. I use 3% but beleave that to be unobtainable. That made the cutoff to be 0.7% to get expansion to 16 PSIA from 1000 PSIA. The three stage compound with 5% clearance and 29.6% cutoff, 3 low expansion of 3.03, to expand to 16 PSIA from 1480 PSIA. I then figured the relative volume of the single sage to the compound first stage so they would have equal quanties of inlet steam. Figured the rest of the compound stages and came up with the 1.5 number. It seamed wrong. The overall isentropic expansion is very close on both engine though the compound operates at a higher first stage inlet pressure. That higher pressure allows for isenthalpic pressure drops between stages that do no work. The over all thoritical cycle efficiency is about the same. So equal amounts of steam would produce equal amounts of power.
The thing is you would not be able run a single stage engine with the small clearances and cutoff. At least I think it would be impossable. The small clearance is required to get the compression and expansion ratio. With a high compression engine, the smaller the clearance the more power you can make for a given displacement. The smaller the clearance the longer the cutoff for the same expansion. Longer cutoff, more steam going through the engine. more power.
I used percentages of steam from my steam cycle calculator. But doing it by cutoffs gets close. We have 0.7% cutoff for single expansion to 29.6% cutoff on firstage of compound. To simplify assume all stage to be the same 29.6% Or haveing a 3:1 expansion ratio. The total volume of the compound would volume v1 of first stage plus volume v2 of second stage and v3 volume of thord stage.
Vc = V1+v2+v3 wher v2 = 3 * v1 and v3 = 3*v2 = 9*v1
vc = v1 + 3*v1 + 9*v1 = 13*v1
The volume of the single expansion Vs is figured so that 0.7% of Vs is equal to 29.6 % of v1
0.07 * Vs = 0.296 * V1
Vs = 0.296/0.07 * V1 = 42.286 V1
Vs/Vc = (42.286 V1) / (13*v1) = 3.253
Vs = 3.253 * Vc
Using steam densities and calculated percentages of inlet steam going the rough a compression cycle I calculated it to be a 1.5:1.
The end result. It seams counter intutitive that the compound could be snaller.
Andy
|
Re: Williams ws. Rankin March 09, 2006 06:28PM |
Registered: 19 years ago Posts: 1,771 |
|
Re: Williams ws. Rankin March 09, 2006 07:40PM |
Registered: 19 years ago Posts: 1,268 |
Hi Andy,
The compound being smaller than a single does seem counter intuitive. I would think that the last stage of the compound would be the same size as the single plus an amount to compensate for the free expansions to the inter stage recievers. This for an equal amount of work producing expansions.
The big idea of a compound is though to produce more expansion than a single so the last stage would be bigger.
It's been hard to get a good realistic number for a percentage of clearance. Obviously with delicate machining of pistons and heads a very low clearance volume could be produced but that is not what we are after. What we need is a minimum clearance percentage that is everyday workable and stable with thermal expansions and dynamics.
Diesel engines do go up to a 27/1 compression ratio which is just under 4% clearance. I think that for a roughly square bore to stroke ratio 4% is a workable percentage. Longer stroke engines might do a little better and I would think that 3% clearance should work ok. Anything less even though machinable might be too unstable and vary with engine temperature and speed too much to be useful. Anyone in the know on this minimum clearance percentage business?
The very small cuttoffs are quite doable with my valve gear design as it varies from zero to maximum cutoff, remembering that the valve doesn't open very far here and doesn't need to, just a squirt of steam near TDC. With recompression the valve starts to open before TDC anyway so the important point is when it closes.
The valve gear also has variable advance so the beginning of opening can be be changed.
Question Andy, since you did many calculations on the subject. On recompression to inlet pressure, is it more detrimental to efficiency to undercompress a bit or overcompress? Thinking of like a hundred pounds one way or another.
Thank You -------- Bill G.
The compound being smaller than a single does seem counter intuitive. I would think that the last stage of the compound would be the same size as the single plus an amount to compensate for the free expansions to the inter stage recievers. This for an equal amount of work producing expansions.
The big idea of a compound is though to produce more expansion than a single so the last stage would be bigger.
It's been hard to get a good realistic number for a percentage of clearance. Obviously with delicate machining of pistons and heads a very low clearance volume could be produced but that is not what we are after. What we need is a minimum clearance percentage that is everyday workable and stable with thermal expansions and dynamics.
Diesel engines do go up to a 27/1 compression ratio which is just under 4% clearance. I think that for a roughly square bore to stroke ratio 4% is a workable percentage. Longer stroke engines might do a little better and I would think that 3% clearance should work ok. Anything less even though machinable might be too unstable and vary with engine temperature and speed too much to be useful. Anyone in the know on this minimum clearance percentage business?
The very small cuttoffs are quite doable with my valve gear design as it varies from zero to maximum cutoff, remembering that the valve doesn't open very far here and doesn't need to, just a squirt of steam near TDC. With recompression the valve starts to open before TDC anyway so the important point is when it closes.
The valve gear also has variable advance so the beginning of opening can be be changed.
Question Andy, since you did many calculations on the subject. On recompression to inlet pressure, is it more detrimental to efficiency to undercompress a bit or overcompress? Thinking of like a hundred pounds one way or another.
Thank You -------- Bill G.
|
Re: Williams ws. Rankin March 09, 2006 08:38PM |
Registered: 19 years ago Posts: 1,453 |
Sorry, only registered users may post in this forum.
All files from this thread
