"Jakuba's Compounding article in the Bulletin"
Posted by jakuba
|
Re: "Jakuba's Compounding article in the Bulletin" September 19, 2014 12:42PM |
Bill. Sounds like you are applying philosophical meaning to entropy. " There would then be an increase in entropy during expansion and another increase during compression"
The entropy of the steam is more then like dropping not increasing. The steam being hotter then the cylinder walls during expansion would have heat transfer from the steam to the cylinder. The makes for a total increase in entropy of the system. Not the steam per say.
Entropy is a property of steam. In a single phase state any two thermodynamic properties can determine a state of the substance. entropy and specific volume {or density= 1/(specific volume)} for example. The fundamental properties are pressure, temperature, specific volume (density), Specific enthalpy, Specific internal energy.
The steam property formulation I implemented are the:
IAPWS Formulation 1995 for the Thermodynamic Properties of Ordinary Water Substance for General and Scientific Use.
The formulations are calculated from the "dimensionless Helmholtz free energy". There two parts. The ideal-gas part of the dimensionless Helmholtz free energy is obtained using an 8 term polynomial equation for the specific isobaric heat capacity in the ideal-gas state. The second part that makes it work for a real gas they call the residual part of the dimensionless Helmholtz free energy. a 56 term polynomial. And then the actual proprieties are combinations of the first and second derivatives. When you do the (calculus) derivatives of those you get even more term. The great thing in using them (other then they are the most accurate formulations) is that they cover the entire range. There are no regions using different formulation were on the boundary there are small differences in the two values calculated for a point on the boundary line. My formulation add a few properties for simulating dynamic processes. isotropic expansion and isenthalip expansion. That involved taking some additional derivatives of thees polynomials.
The point I am trying to make is that even with short abbreviated names those formula fill a page. If you wrote out the names wou would have need several pages just to write down a formula.
The additional properties I added were the rate of change of (reduced temperature with respect to density) on with entropy constant and another for enthalpy constant. The formulation are used in a VisSim plugin. VisSim is a simulation program that actually digitally simulates an analog computer.
Time dependent processes are simulated.
For example, the volume change in an engine cylinder is changing with time. Mathematically we can write relation for that change based say on rotation of the crank. We can mathematically write relation for the a pistons position based on tire rotation. We can do with respect to time. So ultimately we can get the rate of change at any moment of the cylinder volume. And with a bit of manipulation get the rate of change of the steam density in the cylinder.
With my adder properties dT/dD|s rate of change of temperature with respect to density. and the rate of change of density with respect to time the product is the rate of change of temperatue with respect to time dT/dt|s = dT/dD|s * dD/dt. Using an intergrater block I get T out and can calculate steam properties varing with time. Ultimately heat transfer can be added. If we know the heat transfer rate we can adjust the rate of change of temperature for heat transfer and simulate steam expansion in a cylinder combining the ideal isotropic expansion adjusting for heat transfer.
I long explanation of how I came to view entropy and enthalpy changes. All the philosophical ^^^^ doesn;t matter. If density is held constant and you increase enthalpy, entropy increases. Or hold pressure constant the same is true. If you transfer heat to the cylander wall enthalpy decreasses and so does entropy. The decrease in entopy would be less then the increase of entropy in the cylinder wall for a net gain in entropy.
The steam density is deterministic in the cylinder during expansion. It is a function of the initial steam mass and the volume. We are not changing the mass of the steam. It contained in the cylinder. The volume is deterministic dependent on the piston position. At any point the density is deturmined by the mase on volume. So take the case the steams entropy increasing. And sense that would also mean the enthalpy would also increase. Inturn that would also mean that the pressure is great and you would be producing more shaft work.
If we look at heat transfer going both ways during expansion. Initially the steam temperature being higher then the cylinder surface temperature and toward the end end of the stroke the cylander surface being hoter then the relative rate of change of entropy to enthalpy idecreasses with temperature. So in that case entropy may be greater than it initial entry state,
The state of the steam is not just dependent on one process but it's total environment. Below are some expansion charts generated in VisSim.
The Red lines are constant entropy plots. Red are the isentropic expansion combined with a constant rate of temperature added to that of the isentropic.
The very upper left the rate of change of reduced density reduced density is density/(critical point density)
Reduced Temperature is the (critical point temperature divided by the absolout temperature) The IAPWS formulation take as inputs reduced (unit-less) values. My steam property blocks take the reduced unit-less values. It avoids unnecessary conversion calculations on every iteration of the simulation. I provide conversion blocks that do the conversion needed for input parameters and display or ploting. There are a number of state point blocks taking combintions of inputs. The two used in that sim were one block taking reduced temperature and reduced pressure to calculate initial state point. The other two blocks take reduced temperature and reduced density as inputs. The pressure-temperature block gets it's inputs from conversion blocks supplied in the plug-in The initial values are 850 PSI and 1100F. The 1/S blocks are integrator blocks. In this case they are reset integrator blocks that take their value/initial state from their lower input when the midel b input is greater then 0. I use a delay block set the two intergrator to initial reduced temperature and densities from the pressure block. After that they are generating their output from the rate of change input opper input. Inputs are on the left of a block outputs on the right. The rate of change of reduced density Dd is a constant -4. The rate of chage of temperature is being calculated from the rate of change of temperature with respect to density outputs from the temperature density block. the intergrato form a feed back loop calculating their output based on the curent state of the temperature density state point block. One is simply a constant entropy. producing the red plots. The other is adding a temperature increase to the constant isentropic temperature change to produce the blue plots. The upper plot is pressure . The lower temperature and the middle intropies. You can see the red entropy line is flat. It is constant to at least 8 decmal places.
Isentropic expansion does continue into the saturated mixture region. One of the things I havn't gotten working in the dT/dD|s property block. Expanding into the wet steam does not result in water droplets right away. That takes time for droplets to form. It stays a gas for some time and is refered to a subcooled gas state. You allso have a supperheated liquid states. These states are covered by the IAPWS 95 formulation I implemented. but I can find info on the tome factor. it does depend on the saturated quality value. The higher the liquid content the faster the change from a subcool gas to forming water droplets. Compression would still be along an isotropic line. The thermodynamic processes do not care the mater state of a substance. At least that is the way I understand it to work. Thermodynamics is really about energy and the conservation of energy law applies no matter the phisical state of a substance.
The entropy of the steam is more then like dropping not increasing. The steam being hotter then the cylinder walls during expansion would have heat transfer from the steam to the cylinder. The makes for a total increase in entropy of the system. Not the steam per say.
Entropy is a property of steam. In a single phase state any two thermodynamic properties can determine a state of the substance. entropy and specific volume {or density= 1/(specific volume)} for example. The fundamental properties are pressure, temperature, specific volume (density), Specific enthalpy, Specific internal energy.
The steam property formulation I implemented are the:
IAPWS Formulation 1995 for the Thermodynamic Properties of Ordinary Water Substance for General and Scientific Use.
The formulations are calculated from the "dimensionless Helmholtz free energy". There two parts. The ideal-gas part of the dimensionless Helmholtz free energy is obtained using an 8 term polynomial equation for the specific isobaric heat capacity in the ideal-gas state. The second part that makes it work for a real gas they call the residual part of the dimensionless Helmholtz free energy. a 56 term polynomial. And then the actual proprieties are combinations of the first and second derivatives. When you do the (calculus) derivatives of those you get even more term. The great thing in using them (other then they are the most accurate formulations) is that they cover the entire range. There are no regions using different formulation were on the boundary there are small differences in the two values calculated for a point on the boundary line. My formulation add a few properties for simulating dynamic processes. isotropic expansion and isenthalip expansion. That involved taking some additional derivatives of thees polynomials.
The point I am trying to make is that even with short abbreviated names those formula fill a page. If you wrote out the names wou would have need several pages just to write down a formula.
The additional properties I added were the rate of change of (reduced temperature with respect to density) on with entropy constant and another for enthalpy constant. The formulation are used in a VisSim plugin. VisSim is a simulation program that actually digitally simulates an analog computer.
Time dependent processes are simulated.
For example, the volume change in an engine cylinder is changing with time. Mathematically we can write relation for that change based say on rotation of the crank. We can mathematically write relation for the a pistons position based on tire rotation. We can do with respect to time. So ultimately we can get the rate of change at any moment of the cylinder volume. And with a bit of manipulation get the rate of change of the steam density in the cylinder.
With my adder properties dT/dD|s rate of change of temperature with respect to density. and the rate of change of density with respect to time the product is the rate of change of temperatue with respect to time dT/dt|s = dT/dD|s * dD/dt. Using an intergrater block I get T out and can calculate steam properties varing with time. Ultimately heat transfer can be added. If we know the heat transfer rate we can adjust the rate of change of temperature for heat transfer and simulate steam expansion in a cylinder combining the ideal isotropic expansion adjusting for heat transfer.
I long explanation of how I came to view entropy and enthalpy changes. All the philosophical ^^^^ doesn;t matter. If density is held constant and you increase enthalpy, entropy increases. Or hold pressure constant the same is true. If you transfer heat to the cylander wall enthalpy decreasses and so does entropy. The decrease in entopy would be less then the increase of entropy in the cylinder wall for a net gain in entropy.
The steam density is deterministic in the cylinder during expansion. It is a function of the initial steam mass and the volume. We are not changing the mass of the steam. It contained in the cylinder. The volume is deterministic dependent on the piston position. At any point the density is deturmined by the mase on volume. So take the case the steams entropy increasing. And sense that would also mean the enthalpy would also increase. Inturn that would also mean that the pressure is great and you would be producing more shaft work.
If we look at heat transfer going both ways during expansion. Initially the steam temperature being higher then the cylinder surface temperature and toward the end end of the stroke the cylander surface being hoter then the relative rate of change of entropy to enthalpy idecreasses with temperature. So in that case entropy may be greater than it initial entry state,
The state of the steam is not just dependent on one process but it's total environment. Below are some expansion charts generated in VisSim.
The Red lines are constant entropy plots. Red are the isentropic expansion combined with a constant rate of temperature added to that of the isentropic.
The very upper left the rate of change of reduced density reduced density is density/(critical point density)
Reduced Temperature is the (critical point temperature divided by the absolout temperature) The IAPWS formulation take as inputs reduced (unit-less) values. My steam property blocks take the reduced unit-less values. It avoids unnecessary conversion calculations on every iteration of the simulation. I provide conversion blocks that do the conversion needed for input parameters and display or ploting. There are a number of state point blocks taking combintions of inputs. The two used in that sim were one block taking reduced temperature and reduced pressure to calculate initial state point. The other two blocks take reduced temperature and reduced density as inputs. The pressure-temperature block gets it's inputs from conversion blocks supplied in the plug-in The initial values are 850 PSI and 1100F. The 1/S blocks are integrator blocks. In this case they are reset integrator blocks that take their value/initial state from their lower input when the midel b input is greater then 0. I use a delay block set the two intergrator to initial reduced temperature and densities from the pressure block. After that they are generating their output from the rate of change input opper input. Inputs are on the left of a block outputs on the right. The rate of change of reduced density Dd is a constant -4. The rate of chage of temperature is being calculated from the rate of change of temperature with respect to density outputs from the temperature density block. the intergrato form a feed back loop calculating their output based on the curent state of the temperature density state point block. One is simply a constant entropy. producing the red plots. The other is adding a temperature increase to the constant isentropic temperature change to produce the blue plots. The upper plot is pressure . The lower temperature and the middle intropies. You can see the red entropy line is flat. It is constant to at least 8 decmal places.
Isentropic expansion does continue into the saturated mixture region. One of the things I havn't gotten working in the dT/dD|s property block. Expanding into the wet steam does not result in water droplets right away. That takes time for droplets to form. It stays a gas for some time and is refered to a subcooled gas state. You allso have a supperheated liquid states. These states are covered by the IAPWS 95 formulation I implemented. but I can find info on the tome factor. it does depend on the saturated quality value. The higher the liquid content the faster the change from a subcool gas to forming water droplets. Compression would still be along an isotropic line. The thermodynamic processes do not care the mater state of a substance. At least that is the way I understand it to work. Thermodynamics is really about energy and the conservation of energy law applies no matter the phisical state of a substance.
|
Re: "Jakuba's Compounding article in the Bulletin" October 02, 2014 07:43AM |
Hi Bill,
Lack of publicly-available independently-verified test data on any proposed high-efficiency steam-power system can not only be "frustrating", but can also lead to a lack of investment in such systems by folks with the means and desire to invest.
The foregoing comments, of course, are in absolutely no way intended as a comment on any actual proposed steam-power system, or on any actual potential investor in steam-power development, regardless of any "read between the lines" interpretation which somebody might apply to them.
It is my hope that the foregoing meticulously-pruned verbiage will spare me from "being censored into Official Total Non-Existence" by various alleged/self-appointed Internetical Authorities.
That being said, I believe that there may be some commercial potential in a multiple-expansion steam automobile engine, in which cutoff/compresssion are continuously controlled throughout the load/speed range of a real-world automotive drive cycle, to give equal work by each stage. This would allow the classic "smooth-running, direct-drive" operation which steam car drivers prefer, with a much higher potential expansion ratio than in Stanley and similar engines. As was previously mentioned by another poster, the Woolf-compound 4-cylinder/2-stage approach embodied in the Doble "E" has this potential -- but the valve timing of the 2 stages needs to be independently-controlled, by a sophisticated algorithm, to optimize efficiency at varying load/speed levels. That was not done in the original Doble E.
Note that a 4-cyl Doble E could theoretically deliver 4:1 expansion ratios with the same cutoffs & smooth-running as a 2:1 Stanley, at low outputs. At higher outputs/rpms, the expansion ratio could increase to 8:1 or much higher, especially considering the flow-limitation effects which kick in at higher rpm and increase the "effective expansion ratio" above that which is simplistically predicted from valve open/close timing analysis. However, optimizing efficiency requires completely-independent valve-timing control of both stages, balancing smoothness and efficiency, which the Doble E did not have.
Y'all e-mail Mr. Pete, and please do visit me "deep in the heart of Texas" in the near future. In the past few days, my "gotta-do" SoCal to Austin-area TX trans-dimensional transition has been accelerating at a nightmarishly fast pace. It's freakin' me out, man! No BS; please drop in and hang out in my TX joint in a free private guest room, with free use of tons of machine tools & garage/assembly space.
Texas Pete
Lack of publicly-available independently-verified test data on any proposed high-efficiency steam-power system can not only be "frustrating", but can also lead to a lack of investment in such systems by folks with the means and desire to invest.
The foregoing comments, of course, are in absolutely no way intended as a comment on any actual proposed steam-power system, or on any actual potential investor in steam-power development, regardless of any "read between the lines" interpretation which somebody might apply to them.
It is my hope that the foregoing meticulously-pruned verbiage will spare me from "being censored into Official Total Non-Existence" by various alleged/self-appointed Internetical Authorities.
That being said, I believe that there may be some commercial potential in a multiple-expansion steam automobile engine, in which cutoff/compresssion are continuously controlled throughout the load/speed range of a real-world automotive drive cycle, to give equal work by each stage. This would allow the classic "smooth-running, direct-drive" operation which steam car drivers prefer, with a much higher potential expansion ratio than in Stanley and similar engines. As was previously mentioned by another poster, the Woolf-compound 4-cylinder/2-stage approach embodied in the Doble "E" has this potential -- but the valve timing of the 2 stages needs to be independently-controlled, by a sophisticated algorithm, to optimize efficiency at varying load/speed levels. That was not done in the original Doble E.
Note that a 4-cyl Doble E could theoretically deliver 4:1 expansion ratios with the same cutoffs & smooth-running as a 2:1 Stanley, at low outputs. At higher outputs/rpms, the expansion ratio could increase to 8:1 or much higher, especially considering the flow-limitation effects which kick in at higher rpm and increase the "effective expansion ratio" above that which is simplistically predicted from valve open/close timing analysis. However, optimizing efficiency requires completely-independent valve-timing control of both stages, balancing smoothness and efficiency, which the Doble E did not have.
Y'all e-mail Mr. Pete, and please do visit me "deep in the heart of Texas" in the near future. In the past few days, my "gotta-do" SoCal to Austin-area TX trans-dimensional transition has been accelerating at a nightmarishly fast pace. It's freakin' me out, man! No BS; please drop in and hang out in my TX joint in a free private guest room, with free use of tons of machine tools & garage/assembly space.
Texas Pete
|
Re: "Jakuba's Compounding article in the Bulletin" October 05, 2014 05:58AM |
Bill
With piston operated steam engines the steam should be 200 degrees superheated over the saturated steam pressure your running the engine at for best over all efficiency.
Since this is a steam car site I’ll use Stanley car as an example. Most of the larger cars run 600 psi. saturated steam. At 600 psi saturated steam is around 486 F so you should shoot for 686 to 700F coming from your superheater into your engine. The steam cylinder oil is injected directly into the steam line on most cars just after the top of the steam loop. The steam loop is the steam line coming off the output of the superheater and rising up about to the height of the top sheet of the boiler much like a trap. Personally I like to inject it directly at the engine. Steam cylinder oil is ISO 460 or 600. ISO 460 is good up to around 750 F with out oxygen. There is no oxygen or very little in superheated steam. Most Oxygen and Hydrogen are separated out at the water line of the boiler when pure steam is created. I don’t know all that much about how and why. Some use ISO 460 as high as 1200F but at 1200F it should be ISO 600.
A Stanley oil pump has a 1/8 D piston and about a ½ stroke and pumps oil about every four revolutions of the engine. Hence my comment, pissey ass amount compared with any gas or diesel engine for cylinder lubrication.
Rolly
With piston operated steam engines the steam should be 200 degrees superheated over the saturated steam pressure your running the engine at for best over all efficiency.
Since this is a steam car site I’ll use Stanley car as an example. Most of the larger cars run 600 psi. saturated steam. At 600 psi saturated steam is around 486 F so you should shoot for 686 to 700F coming from your superheater into your engine. The steam cylinder oil is injected directly into the steam line on most cars just after the top of the steam loop. The steam loop is the steam line coming off the output of the superheater and rising up about to the height of the top sheet of the boiler much like a trap. Personally I like to inject it directly at the engine. Steam cylinder oil is ISO 460 or 600. ISO 460 is good up to around 750 F with out oxygen. There is no oxygen or very little in superheated steam. Most Oxygen and Hydrogen are separated out at the water line of the boiler when pure steam is created. I don’t know all that much about how and why. Some use ISO 460 as high as 1200F but at 1200F it should be ISO 600.
A Stanley oil pump has a 1/8 D piston and about a ½ stroke and pumps oil about every four revolutions of the engine. Hence my comment, pissey ass amount compared with any gas or diesel engine for cylinder lubrication.
Rolly
Sorry, only registered users may post in this forum.
