Monotube Flash Boiler Design

[Toymaker]

The sonic velocity V of a gas is determined by the constants R, K, g and the value of its temperature.
That is V= square root (gkRT). Note that this value increases rapidly with T. Seven degrees does not guarantee smooth subsonic flow. Turbulent flow is always a given because the Reynolds number are way above 2000. There are two types of turbine blades and each has a different way of causing a pressure drop. One is impluse and the other is reaction blade. The angles are derived from vector equations and the shape of the blades and the nozzle block is designed specifically for its application. This is a basic thermodynamic consideration. So you will need to consider the gas Temperature and then calculate the velocity through all the passages and if its over the sonic velocity choke flow will result.

First, thanks for the sonic velocity equation,...it's one I haven't seen before. I've found that with a bit of digging, I can usually find a chart or graph showing speed of sound vs temperature at various pressures; the graph below is for R123:


The last sentence under the sub-heading "VI Conclusion" of the attached research paper on finding the ideal taper angle states: The optimum taper angle is 7ᴼ below which there is no flow separation at all but going beyond it gives rise to flow separation.

I've chosen to use a combined impulse-reaction design for my blades, as shown below. The small circles show the dimensions of the smallest passage between blades.

 

[Steamchick]

I see from the graph how you can determine the speed of sound at 183deg.C, and 30MPa. = 500m/s. I guess it is possible to design your nozzles to not exceed that speed of gas where they eject the gas stream at the blades?
If my crude understanding is correct, the gas will slow as it expands, and although the speed of sound will also slow down as the pressure drops, these are inter-related to ensure the gas does not exceed the speed of sound as it passes through the open space to the first blades, after which they will have lost maybe <50% of the kinetic and thermal energy, but will be much slower than the speed of sound at the lower pressure anyway?
I think the tremendous "Roar" from the NASA sized space rockets is because the combusting gases exceed the speed of sound inside the nozzles, as the gas is burning and chemical energy is added during the expansion in the nozzles. In your R123 case, the gas is endothermic, as there is no chemical release of energy involved. So it follows the gas laws.
Is any of that sensible?
K2

 

[Toymaker]

I see from the graph how you can determine the speed of sound at 183deg.C, and 30MPa. = 500m/s. I guess it is possible to design your nozzles to not exceed that speed of gas where they eject the gas stream at the blades?
If my crude understanding is correct, the gas will slow as it expands, and although the speed of sound will also slow down as the pressure drops, these are inter-related to ensure the gas does not exceed the speed of sound as it passes through the open space to the first blades, after which they will have lost maybe <50% of the kinetic and thermal energy, but will be much slower than the speed of sound at the lower pressure anyway?
I think the tremendous "Roar" from the NASA sized space rockets is because the combusting gases exceed the speed of sound inside the nozzles, as the gas is burning and chemical energy is added during the expansion in the nozzles. In your R123 case, the gas is endothermic, as there is no chemical release of energy involved. So it follows the gas laws.
Is any of that sensible?
K2

With the exception of Convergent-Divergent nozzles, flowing gases never exceed local Mach, and as you mentioned above, gas velocity quickly decreases as the gases begin to expand once they leave the nozzle area. There's no need to design blades, stators, or the shape of the turbine walls to prevent gasses from exceeding the speed of sound, because, with the exception of the very short distances within CD nozzles, gas flow never exceeds local Mach.

Looking at the blade-stator drawing above (post#221) the first blade row is pure impulse, but notice how the entrance area of both the stators and the blades is much larger than their exit area; this shape forms a convergent nozzle which acts to increase gas velocity. Notice too the direction of gas flow as it leaves the stators and impacts the blades nearly flat surface at the front of the blades; causing the front 1/3 of the blades to act as impulse blades. Then, as the gases move through the blades the gases are forced through a much smaller area which causes the gases to increase velocity as they exit the blade row, effectively acting as a reaction blade.

 

[Steamchick]

Thanks, Your explanation is a help. I have just been looking at the diffuser angle paper. - briefly.
I have designed venturi and expanding orifi diffuser tubes for inducing air into gas burners, and used 9 deg. tapers which came from a old paper on the subject. I future, I'll reduce that to 7~8 degrees.
I wonder how much of the initial set of blades is reaction (velocity from momentum exchange developing pressure on the back face), and how much "aerodynamic"? - induced pressure drop from the shape induced velocity difference between the leading face and back-face of each blade? (longer path on leading face compared to back-face, like a wing, lowering the pressure on the leading face, increasing the pressure difference across the blade?).
I really don't know what I am talking about, so may have used the wrong words in this...
Cheers. K2

 

[Toymaker]

Thanks, Your explanation is a help. I have just been looking at the diffuser angle paper. - briefly.
I have designed venturi and expanding orifi diffuser tubes for inducing air into gas burners, and used 9 deg. tapers which came from a old paper on the subject. I future, I'll reduce that to 7~8 degrees.
I wonder how much of the initial set of blades is reaction (velocity from momentum exchange developing pressure on the back face), and how much "aerodynamic"? - induced pressure drop from the shape induced velocity difference between the leading face and back-face of each blade? (longer path on leading face compared to back-face, like a wing, lowering the pressure on the leading face, increasing the pressure difference across the blade?).
I really don't know what I am talking about, so may have used the wrong words in this...
Cheers. K2

I think I understand your question; you want to know what percentage of the torque produced by each blade comes from the impact of the gas stream hitting the concave surface vs how much torque comes from each blade acting as an airplane wing and producing lift on the convex surface. Unfortunately, the answer exceeds my knowledge. I will only note that all the text I've read speak only of the impulse forces.

Seems like a good question for Google or ChatGPT

 

[Steamchick]

I'm glad you can make more sense of my question than I can!. That is my question, and a respectable answer.
thanks.
K2

 

[Toymaker]

Return of the boiler's Radiant Disk (with video): The drawing below shows the boiler's original design layout which has the hot exhaust gases from the burner shooting directly onto a hollow disk I call the Radiation Disk; so named because it's location would absorb much of the radiant heat energy from the exhaust flames. Months ago I reluctantly gave up on the radiant disk idea as it proved a nightmare to fabricate.

At some point, I decided to machine the disk from a solid chunk of 6061 aluminum, and bolt on a 1/4" aluminum plate to cover the machined channels. A single high temperature silicon O-ring seals the back plate onto the disk's main body. I wasn't at all certain the aluminum would not melt from being directly in the path of the 2500 C blue flames, even though the disk was being cooled by the water/steam flowing through it, .... but I really wanted to use my radiant disk idea, so I opted to give it a try.

Below photo's don't show the internal M18 x 1.5 threads which the brass connectors thread into, and which the tubes will be brazed onto, as shown in the video. The short copper tubes seen in the below video are use for hydrostatic testing and will be brazed into the final boiler assembly.




I was pleasantly surprised when the aluminum didn't melt

Next step: Hydrostatically pressure test the disk.

 

[NapierDeltic]

It resembles somehow a cylinder head ...
I am pretty sure you have made the dimensioning calculations and considered safety factors, but because it looks to me the most dangerous part of the vapour circuit in case of malfunction, I consider appropriate to ask for it.
Good luck!

 

[Toymaker]

It resembles somehow a cylinder head ...
I am pretty sure you have made the dimensioning calculations and considered safety factors, but because it looks to me the most dangerous part of the vapour circuit in case of malfunction, I consider appropriate to ask for it.
Good luck!

Thanks for the Good Luck,...I hope I don't need too much of it

The resemblance to a cylinder head is intentional. The fins on an air cooled cylinder head are there to dissipate heat from the engine assembly, while the fins on my disk are intended to move heat into the working fluid inside the disk; in both cases the fins increase the surface area in contact with the outside gasses,... which is air for an IC engine and combustion gases inside the boiler.

My design approach was to design for a typical pressure vessel. First calculate total force on the cover plate: 4" diameter exposed to vapor pressure yields 12.56 sq in area. Area x 1000 psi = 12,566 lbs force on the cover plate, which is actually twice what I will need since my max operating pressure will be just over 500 psi. One M4 x 0.7 stainless screw has 1,380 lbf yield strength times 20 screws = 27,616, which is just over twice what I need. Finding the stripping strength for the aluminum threads is more challenging; rule-of-thumb recommendations are 1.5 to 2.5 times the diameter of the screw. I have 14mm of thread engagement on the 4mm screws, or 3.5 times the diameter; which should be more than enough.

Just to make sure, I will hydrostatically test the disk before I braze it into the boiler tubes.

 

[Steamchick]

I worry a bit about your "factor of safety of 2" - or 4 or whatever - when you decide how strong things should be. In Western Regulations, for compressed gases and boilers it is common to design to a factor of safety of 8 as a minimum. Old designs may have used FOS = 6, but only applicable where they have been proven failure-free for 35 years or more.
But the proof is in the final product, so I hope you are right....
K2

 

[Toymaker]

When nothing goes wrong, watching a hydrostatic test is a bit like watching paint dry,...so, no video, just a pic of the test set-up, which I performed outside in the yard; in case something leaked, I don't get hydraulic oil all over the shop. The disk is placed on a sheet of white paper to better see any leaks, should they occur.

Pressure was taken up to 1000 psi in steps of 50 psi. The two threaded fittings on the disk are sealed with Permatex Ultra Copper silicon sealant, which has a max temperature limit of 371 C (700 F),...well above my planned 200 C max limit. There were no surprise leaks,... all went as planned.

Next steps: braze the disk onto the boiler tubes, followed by a final hydrostatic pressure test of the boiler tubes & disk combination.

 

[Toymaker]

I worry a bit about your "factor of safety of 2" - or 4 or whatever - when you decide how strong things should be. In Western Regulations, for compressed gases and boilers it is common to design to a factor of safety of 8 as a minimum. Old designs may have used FOS = 6, but only applicable where they have been proven failure-free for 35 years or more.
But the proof is in the final product, so I hope you are right....
K2

The pressure vessels many of us fly in on long trips (aka: commercial aircraft) are designed using a FoS of 2. Non-pressurized aircraft typically use FoS of 1.5, and the landing gear on pretty much all aircraft use a FoS 1.25.

I'm 100% comfortable using an FoS of 2,...IMHO, more than 2 is overkill.

 

[Steamchick]

I blindly accept you are right for pressures in aircraft.... but for UK, Europe, USA, Australia, etc. there are regulations for pressure vessels that "in their professional wisdom" use higher Factors of Safety for stressed components of pressure vessels. So to comply with local testing and design requirements I must follow those Regulations. As long as your boiler does not fail you can call it a success. I hope you are right. But please use Irish safety: "Be sure to be sure".
K2

 

[Toymaker]

I blindly accept you are right for pressures in aircraft.... but for UK, Europe, USA, Australia, etc. there are regulations for pressure vessels that "in their professional wisdom" use higher Factors of Safety for stressed components of pressure vessels. So to comply with local testing and design requirements I must follow those Regulations. As long as your boiler does not fail you can call it a success. I hope you are right. But please use Irish safety: "Be sure to be sure".
K2

Buddha told his followers to accept conditionally what others presented as facts,...and only after you've had a chance to test the presented "facts" and determine for yourself their accuracy, should you accept them as factual, and accurate.

Therefore, please don't blindly accept anything I post until you've had a chance to check for yourself my accuracy,...I do my best to never post false information, but I do sometimes get it wrong. If you Google a phrase like this one, "what is faa fos for aircraft", or a similar phrase, you'll get lots of relevant info.

I honestly expected someone to ask, "why is the Allen wrench in the photo" ? The Allen wrench is there "to be sure". For me, the highest risk in my design were the aluminum threads holding those 20 M4 screws. The math told me the stainless screws are unlikely to fail, but I wasn't so confident the screws would not shear off the aluminum threads and pull out. So, after the initial pressure of 500 psi I released the pressure and checked the tightness of all 20 screws; a loose screw would have indicated the aluminum threads were failing. I followed this procedure for each 50 psi increase, up to 1000 psi. The screws held, without loosening.

 

[Steamchick]

I like your thoroughness. And will accept "blindly" what you tell me about the FAA rules, as I trust them every time I fly...
Buddha was a great philosopher. Wish he had been an Engineer though.... just think of what he could have designed? Could he have imagined something as far reaching as the WWW?
Back to "boilers".... I'm sure you will have checked for any temperature issues with the screws and aluminium having such different coefficients of thermal expansion. My understanding of "1950s engines" where "traditional" designs replaced cast iron cylinder heads with aluminium, and later design work on 1980s car engines of cast iron blocks and aluminium heads and Aluminium blocks with aluminium heads - all with steel bolts - and the calculations of increased stresses in the aluminium from the expansion of hot aluminium versus the expansion of the steel bolts all leads me to advise you have done the sums to ensure the aluminium does not plastically distort (locally compress) when at the elevated temperature. In car engines et. it is common to use long bolts, not short ones, to permit the increased stress to be limited to less than the yield stress of the aluminium.

I hope this explanation is adequate for you to understand something that I am a bit vague about, as it is >30 years since I was involved in the calcs.?
I think it is a combination of the stress of tightening the bolt => extension of the bolt, plus the stress of: the expansion of the aluminium to max. working temperature - the expansion of the steel to max. working temperature. I experienced this with the bolt heads having large and thick washers beneath the bolt heads/nuts to prevent cast aluminium cylinder heads from being locally distorted - to the extent that the assembled bolting pressure as lost on head gaskets at cold starting such that the head gaskets would blow. All a part of engine design - especially for modern pressed steel shim type head gaskets that could be over compressed during heating and fail to recover for cold starting.
In your case, it may be necessary to provide large washers beneath the Allen screw heads of the bolts to prevent local compression (a later leakage due to loss of contact pressure at the joint) of the aluminium?
Regards,
K2

 

[Toymaker]

I like your thoroughness. And will accept "blindly" what you tell me about the FAA rules, as I trust them every time I fly...
Buddha was a great philosopher. Wish he had been an Engineer though.... just think of what he could have designed? Could he have imagined something as far reaching as the WWW?
Back to "boilers".... I'm sure you will have checked for any temperature issues with the screws and aluminium having such different coefficients of thermal expansion. My understanding of "1950s engines" where "traditional" designs replaced cast iron cylinder heads with aluminium, and later design work on 1980s car engines of cast iron blocks and aluminium heads and Aluminium blocks with aluminium heads - all with steel bolts - and the calculations of increased stresses in the aluminium from the expansion of hot aluminium versus the expansion of the steel bolts all leads me to advise you have done the sums to ensure the aluminium does not plastically distort (locally compress) when at the elevated temperature. In car engines et. it is common to use long bolts, not short ones, to permit the increased stress to be limited to less than the yield stress of the aluminium.

I hope this explanation is adequate for you to understand something that I am a bit vague about, as it is >30 years since I was involved in the calcs.?
I think it is a combination of the stress of tightening the bolt => extension of the bolt, plus the stress of: the expansion of the aluminium to max. working temperature - the expansion of the steel to max. working temperature. I experienced this with the bolt heads having large and thick washers beneath the bolt heads/nuts to prevent cast aluminium cylinder heads from being locally distorted - to the extent that the assembled bolting pressure as lost on head gaskets at cold starting such that the head gaskets would blow. All a part of engine design - especially for modern pressed steel shim type head gaskets that could be over compressed during heating and fail to recover for cold starting.
In your case, it may be necessary to provide large washers beneath the Allen screw heads of the bolts to prevent local compression (a later leakage due to loss of contact pressure at the joint) of the aluminium?
Regards,
K2

My solution to avoid leaks due to thermal expansion-contraction cycling was to use high temperature O-rings on all the flat surfaces; I'm currently using red silicone, (see post #227) which is good to about 230 C, but I will replace those with FFKM O-rings (green) whenever they arrive in my mail box. FFKM O-rings are not available locally, so I must order them on-line, but they're rated for 300 C, which is well above the 220 C absolute max I plan to use.

I like your idea of using flat washers under the cap-head screws to avoid deforming the aluminum,...I need to look for my M4 stainless washers

 

[Toymaker]

Brazed in the radiation disc into the boiler tubes followed by an uneventful final hydrostatic test to 1000 psi. Next steps: insert the tube assembly and burner assembly into the boiler housing. Rigidly mount the ignitor box. Build framework to restrain boiler on test stand.

 

[Richard Hed]

Brazed in the radiation disc into the boiler tubes followed by an uneventful final hydrostatic test to 1000 psi. Next steps: insert the tube assembly and burner assembly into the boiler housing. Rigidly mount the ignitor box. Build framework to restrain boiler on test stand.
View attachment 150792 View attachment 150793

what I like about water-tube type boilers over fire-tube boilers has several factors: they come up to pressure faster, they are far safer because of when reaching their limit failure point, the fire tube tends to have catastrophic explosions releasing the energy in an instant. But the water tube type, when reaching failure generally fail at a pin point or a very small split which releases the energy more controlled (not necessarily without danger, but it doesn't tend to take out a city block).

 

[Steamchick]

Providing the pump can maintain pressure AND FLOW they are practical and safe. But "those that have made them" tell me that it is a difficult balance to achieve... and maintain. "Fast Hydo" boats tend to have a direct displacement engine driven "coolant" (boiler feed) pump, and fuel pump, so feed flow/volume is directly a function of engine speed/steam-demand. Engine stops water stops. Usually when the fuel tank is emptied! - So no melt down.
I understand (but have never been there!) that a control factor needs to be max pressure, at max fire and feed. If the pressure rises above the pump's ability, then the pump cannot feed "coolant" to the boiler, so it rapidly overheats to failure. So a Pressure relief Safety Valve needs to be incorporated. (In Toymakers refrigerant gas application the exhaust from the Safety Valve must be through a separate condenser to the "cold" liquid reservoir - Large enough to take full flow of gas from the Safety valve. But I am sure he will have this planned...).
The Monotube boiler application here is for a continuous use boiler, so should not be a problem (once feed/demand is balanced) so Toymaker has some interesting engineering here, controlling the pump flow versus fire. I guess some feedback loop between pressure and burner fuel supply? - Perhaps a directly coupled pressure driven fuel valve that shuts off the fuel if pressure is exceeded? Or a trip that shuts off fuel if the Safety Valve lifts? - But that is the "control system" that may be a different thread to this one about the boiler.
We will learn in due course.
K2

 

[Toymaker]

Providing the pump can maintain pressure AND FLOW they are practical and safe. But "those that have made them" tell me that it is a difficult balance to achieve... and maintain. "Fast Hydo" boats tend to have a direct displacement engine driven "coolant" (boiler feed) pump, and fuel pump, so feed flow/volume is directly a function of engine speed/steam-demand. Engine stops water stops. Usually when the fuel tank is emptied! - So no melt down.
I understand (but have never been there!) that a control factor needs to be max pressure, at max fire and feed. If the pressure rises above the pump's ability, then the pump cannot feed "coolant" to the boiler, so it rapidly overheats to failure. So a Pressure relief Safety Valve needs to be incorporated. (In Toymakers refrigerant gas application the exhaust from the Safety Valve must be through a separate condenser to the "cold" liquid reservoir - Large enough to take full flow of gas from the Safety valve. But I am sure he will have this planned...).
The Monotube boiler application here is for a continuous use boiler, so should not be a problem (once feed/demand is balanced) so Toymaker has some interesting engineering here, controlling the pump flow versus fire. I guess some feedback loop between pressure and burner fuel supply? - Perhaps a directly coupled pressure driven fuel valve that shuts off the fuel if pressure is exceeded? Or a trip that shuts off fuel if the Safety Valve lifts? - But that is the "control system" that may be a different thread to this one about the boiler.
We will learn in due course.
K2

I have the greatest admiration and respect for those whom have built a montotube boiler and control it through mechanical means,...way too big a challenge for me to even consider.

However, controlling most any machine with a micro-controller (aka computer) and the needed sensors vastly simplifies things. My burner/boiler assembly will be fully controlled by an ECU (Engine Control Unit, aka a computer) and will have a pressure sensor on the input into the boiler, (aka, output from the feed pump) and a pressure sensor on the boilers steam output. If the water reserve tank were ever to run out of water, the first indication would be zero feed pump pressure, resulting in the boiler input sensor sending a low pressure signal to the ECU which would instantly shut off the burner, preventing any damage to the boiler tubes. The ECU examines all the various pressure and temperature sensors dozens of times per second, reacting far faster than the boiler is able to exceed pressure or temperature limits.

The ECU has full control over feed pump, fuel flow, and burner air flow; when the human operator requests more power, the ECU increases fuel and air flow and increases feed pump pressure while also watching steam output pressure and temperature. When less power is requested, the ECU reduces fuel, burner air flow and feed pump pressure, but not so fast that the input pressure becomes less than the output pressure.

I talk about the ECU extensively in this thread: Ambitious ORC Turbine

I've posted several videos to YouTube showing the burner under ECU control;
ECU controlled burner
Night Burn under ECU control

K2, why do you believe I need to route the "steam" from the safety valve through a separate condenser? Assuming the normal condenser can handle the additional load, why not use the normal condenser?