Rahul -
Looking at your dimensioned sketch, the length proportions look pretty good to me, at least as a starting point. The only big criticism I have is that the intake ID is very small in proportion to the tailpipe ID. I don't think this will keep the engine from running, but it will have at least two effects: It will make tuning the lengths more critical because the intake will resonate very sharply at its own characteristic frequency; and, it will probably limit the engine's power - like a long-distance runner breathing through a straw. The air speed through such a small intake will be pretty high, but I don't think that in itself will be a problem for an engine this size.
I'll try to set it up in UFLOW tomorrow at lunch time and then we'll get to see what it really looks like.
L Cottrill
~~~~First Valveless!!~~~~
Moderator: Mike Everman
-
larry cottrill
- Posts: 4140
- Joined: Sun Oct 05, 2003 1:17 am
- Antipspambot question: 0
- Location: Mingo, Iowa USA
- Contact:
New Engine
Last edited by larry cottrill on Tue Feb 20, 2007 7:42 pm, edited 1 time in total.
-
_Rahul_
- Posts: 129
- Joined: Sat Jul 15, 2006 2:47 am
- Antipspambot question: 0
- Location: India
- Contact:
Finally planned to use an electric gas lighter circuit for the sparking system. Its cheap, not too cheap though, but fits somewhat into my budget. It runs on 5 volts and sparks ain't too hot. But its meant for lighting LPG stoves so it must ignite the LPG into the PJ as well.
Just a bit worried about the dia of the intake pipe .. waiting for the Uflows' output .. Larry!
Looking at your dimensioned sketch, the length proportions look pretty good to me, at least as a starting point. The only big criticism I have is that the intake ID is very small in proportion to the tailpipe ID. I don't think this will keep the engine from running, but it will have at least two effects: It will make tuning the lengths more critical because the intake will resonate very sharply at its own characteristic frequency; and, it will probably limit the engine's power - like a long-distance runner breathing through a straw. The air speed through such a small intake will be pretty high, but I don't think that in itself sill be a problem for an engine this size.
I'll try to set it up in UFLOW tomorrow at lunch time and then we'll get to see what it really looks like.
L Cottrill
Just a bit worried about the dia of the intake pipe .. waiting for the Uflows' output .. Larry!
[url=callto://ralph_10019]
[/url]
[/url]-
larry cottrill
- Posts: 4140
- Joined: Sun Oct 05, 2003 1:17 am
- Antipspambot question: 0
- Location: Mingo, Iowa USA
- Contact:
UFLOW1D Analysis - Rahul's FWE
Rahul -
Here is a preliminary analysis of your engine, based on your dimensioned sketch. Please review the drawing below ("Physical Dimensions (original)"), and determine whether the dimensions and angles are close to what you really have. This is meant to be a scale drawing, and should look just like the proportions of your real engine. The most important dimensions are shown in the next drawing ("Dimensions for Analysis"), and I simply got them by scaling from the first drawing. The third drawing ("Unfolded Layout for UFLOW1D Analysis") shows how this layout was 'unfolded' as a linear engine model that UFLOW1D can handle - I turn this around end-to-end as I enter the UFLOW data, just to make it easier to do changes in the intake length for tuning the model.
Some explanation of how I did this analysis:
I assume that we want (and can get) the velocity node (pressure antinode) to be at the front plate, and this is the point around which I unfold the intake/transition zone. I consider everything from the node out to the near end of the intake pipe to be the 'transition zone'. On a folded engine like this, I assume that all of the transition zone is part of the chamber and has the full chamber temperature. This is why the 2000 degK chamber temperature shows up in the transition cone in the UFLOW layout drawing. Thus, I assume that there will be a sharp break in temperature between the transition zone and the intake pipe, as shown. On the other hand, I assume that the tailpipe will gradually decrease in temp as we move rearward from the chamber.
I first do a run of the engine as a whole, to get a value for operating frequency and to make sure that the velocity node is right where we want it. Then, I duplicate the geometry and temps of the transition zone and intake as a quarter-wave 'flask' (as I have described before), and I make whatever minor adjustments are needed to the intake length to get it to oscillate as nearly as possible to the whole-pipe frequency. Then, I transition the new intake dimension back to the whole pipe model and run again, noting the change in frequency. I go back and forth a few times until the flask and the whole pipe model agree very closely in terms of dimensions and frequency. This assures that the front end of the unfolded model is really in resonance with the rear section.
It turned out that your original dimensions are very close to being in tune for this pipe - I had to add a whopping 6 mm to the intake length to get the engine tuned! That is so small a change that my errors in estimating temperatures could cover the difference, which means you might have an engine that will run right away!!! I certainly would NOT modify the lengths until you have seriously tried to get it running (playing with fuel spout location or whatever). I would start out with a simple straight fuel pipe spouting right inside the intake rim. You can find a deeper, more efficient location later experimentally.
The UFLOW1D printouts are all run with my final dimensions (the 6 mm added to the intake length). The first graphic is what I used for the starting condition: low pressure spread throughout the pipe, ready to make the engine draw breath. Interestingly, note that UFLOW has provided a datum mark at the .685 metre point, only 1mm from the desired velocity node / pressure antinode. The density curve in the first graphic shows how low pressure and high gas temps impact the air density (all below 1.0, which is normal air).
Once UFLOW is run, we can walk through the first few cycles. The second grahic shows the approximate moment of maximum pressurization of the chamber due to Kadenacy action. This is supposed to represent the condition just before the explosion takes place. Note at this moment in the cycle that the velocity and massflow curves are fairly low in value - a lot of air has poured into the pipe from both the tail and the intake inlet, but most of it has stopped moving as it has charged the pipe with pressure. Note that the density is high only at the two ends, where hot air has been more or less fully replaced - the main part of the engine still contains little more than hot gas residue.
The third graphic shows the 'running out' of our pipe OVER TIME (0.319 second). Remember that we've done nothing in UFLOW to simulate an air/fuel explosion, so we're just seeing how the ebb and flow of air and changes of pressure take place over time, with no energy addition. Note that because of this, there is inevitably a 'running down' of energy over the first few cycles. That is normal for this kind of UFLOW run - what we want to see is that the pipe 'runs out' well with as little degradation as possible per cycle. This picture is about as good as a relatively short straight pipe engine gets, in my experience. The three lines represent three different points in the pipe: Dark blue = just inside the tailpipe exit; Yellow = at the velocity node (the front wall); and Cyan = just inside the intake inlet. In the Pressure graph, we see the full range of pressure development at the front wall - this is the pressure antinode, so this is the greatest pressure swing we'll see anywhere in the pipe. Note how flat the yellow line is in the Mach Number and Mass Flow graphs - we really ARE at the velocity node! Note the phase difference in the blue and cyan lines in those two graphs - above zero is gas movement from left to right, below zero from right to left. Note that the intake Mach Number is very close to the tailpipe speed in the same cycle - this is somewhat unusual, but it's just the result of the small intake pipe diameter - see how small the intake Mass Flow is, compared to the tailpipe Mass Flow! This combination is typical of small intake engines.
As I said, this shows really good operation for a relatively short straight pipe. It is very similar to my original Short Lady in proportion (even the smaller than ideal intake - I only had about 16 or 17mm ID), but is a little larger engine overall. There is every reason to think this engine should run well with minimal tweaking of the lengths (and PERHAPS with no changes at all!).
Let me know if you find any error in the dimensions I used for this analysis. As I said earlier, the critical dimensions are the ones shown on the drawing labeled 'Dimensions for Analysis' - those are the ones I need to correct if they happen to be off by more than a couple of mm.
Good Luck testing your engine!
L Cottrill
Here is a preliminary analysis of your engine, based on your dimensioned sketch. Please review the drawing below ("Physical Dimensions (original)"), and determine whether the dimensions and angles are close to what you really have. This is meant to be a scale drawing, and should look just like the proportions of your real engine. The most important dimensions are shown in the next drawing ("Dimensions for Analysis"), and I simply got them by scaling from the first drawing. The third drawing ("Unfolded Layout for UFLOW1D Analysis") shows how this layout was 'unfolded' as a linear engine model that UFLOW1D can handle - I turn this around end-to-end as I enter the UFLOW data, just to make it easier to do changes in the intake length for tuning the model.
Some explanation of how I did this analysis:
I assume that we want (and can get) the velocity node (pressure antinode) to be at the front plate, and this is the point around which I unfold the intake/transition zone. I consider everything from the node out to the near end of the intake pipe to be the 'transition zone'. On a folded engine like this, I assume that all of the transition zone is part of the chamber and has the full chamber temperature. This is why the 2000 degK chamber temperature shows up in the transition cone in the UFLOW layout drawing. Thus, I assume that there will be a sharp break in temperature between the transition zone and the intake pipe, as shown. On the other hand, I assume that the tailpipe will gradually decrease in temp as we move rearward from the chamber.
I first do a run of the engine as a whole, to get a value for operating frequency and to make sure that the velocity node is right where we want it. Then, I duplicate the geometry and temps of the transition zone and intake as a quarter-wave 'flask' (as I have described before), and I make whatever minor adjustments are needed to the intake length to get it to oscillate as nearly as possible to the whole-pipe frequency. Then, I transition the new intake dimension back to the whole pipe model and run again, noting the change in frequency. I go back and forth a few times until the flask and the whole pipe model agree very closely in terms of dimensions and frequency. This assures that the front end of the unfolded model is really in resonance with the rear section.
It turned out that your original dimensions are very close to being in tune for this pipe - I had to add a whopping 6 mm to the intake length to get the engine tuned! That is so small a change that my errors in estimating temperatures could cover the difference, which means you might have an engine that will run right away!!! I certainly would NOT modify the lengths until you have seriously tried to get it running (playing with fuel spout location or whatever). I would start out with a simple straight fuel pipe spouting right inside the intake rim. You can find a deeper, more efficient location later experimentally.
The UFLOW1D printouts are all run with my final dimensions (the 6 mm added to the intake length). The first graphic is what I used for the starting condition: low pressure spread throughout the pipe, ready to make the engine draw breath. Interestingly, note that UFLOW has provided a datum mark at the .685 metre point, only 1mm from the desired velocity node / pressure antinode. The density curve in the first graphic shows how low pressure and high gas temps impact the air density (all below 1.0, which is normal air).
Once UFLOW is run, we can walk through the first few cycles. The second grahic shows the approximate moment of maximum pressurization of the chamber due to Kadenacy action. This is supposed to represent the condition just before the explosion takes place. Note at this moment in the cycle that the velocity and massflow curves are fairly low in value - a lot of air has poured into the pipe from both the tail and the intake inlet, but most of it has stopped moving as it has charged the pipe with pressure. Note that the density is high only at the two ends, where hot air has been more or less fully replaced - the main part of the engine still contains little more than hot gas residue.
The third graphic shows the 'running out' of our pipe OVER TIME (0.319 second). Remember that we've done nothing in UFLOW to simulate an air/fuel explosion, so we're just seeing how the ebb and flow of air and changes of pressure take place over time, with no energy addition. Note that because of this, there is inevitably a 'running down' of energy over the first few cycles. That is normal for this kind of UFLOW run - what we want to see is that the pipe 'runs out' well with as little degradation as possible per cycle. This picture is about as good as a relatively short straight pipe engine gets, in my experience. The three lines represent three different points in the pipe: Dark blue = just inside the tailpipe exit; Yellow = at the velocity node (the front wall); and Cyan = just inside the intake inlet. In the Pressure graph, we see the full range of pressure development at the front wall - this is the pressure antinode, so this is the greatest pressure swing we'll see anywhere in the pipe. Note how flat the yellow line is in the Mach Number and Mass Flow graphs - we really ARE at the velocity node! Note the phase difference in the blue and cyan lines in those two graphs - above zero is gas movement from left to right, below zero from right to left. Note that the intake Mach Number is very close to the tailpipe speed in the same cycle - this is somewhat unusual, but it's just the result of the small intake pipe diameter - see how small the intake Mass Flow is, compared to the tailpipe Mass Flow! This combination is typical of small intake engines.
As I said, this shows really good operation for a relatively short straight pipe. It is very similar to my original Short Lady in proportion (even the smaller than ideal intake - I only had about 16 or 17mm ID), but is a little larger engine overall. There is every reason to think this engine should run well with minimal tweaking of the lengths (and PERHAPS with no changes at all!).
Let me know if you find any error in the dimensions I used for this analysis. As I said earlier, the critical dimensions are the ones shown on the drawing labeled 'Dimensions for Analysis' - those are the ones I need to correct if they happen to be off by more than a couple of mm.
Good Luck testing your engine!
L Cottrill
- Attachments
-
- The FWE 'running out' from the initial low pressure distribution shown above. Graphic Copyright 2007 Larry Cottrill
- UFLOW_Rahuls_FWE_running_out.GIF (33.54 KiB) Viewed 11837 times
-
- Best pressurization of the chamber area, as the inflow of air from the intake and tailpipe grind to a halt. Graphic Copyright 2007 Larry Cottrill
- UFLOW_Rahuls_FWE_best_pressure.GIF (28.47 KiB) Viewed 11837 times
-
- Starting condition for the UFLOW run of Rahul's engine. Low pressure is distributed nonuniformly throughout the pipe - a picture of Kadenacy action after the explosion phase of the cycle. Graphic Copyright 2007 Larry Cottrill
- UFLOW_Rahuls_FWE_starting_cond.GIF (27.55 KiB) Viewed 11837 times
-
- Rahuls_new_FWE.GIF
- Rahul's new FWE, drawn to scale from his dimensioned sketch given earlier, and also showing how the 'unfolded' linear model is developed for UFLOW1D analysis. Drawing Copyright 2007 Larry Cottrill
- (12.76 KiB) Downloaded 179 times
-
_Rahul_
- Posts: 129
- Joined: Sat Jul 15, 2006 2:47 am
- Antipspambot question: 0
- Location: India
- Contact:
Thanx for the BIG help Larry!
All dimensions are almost correct (+- 1mm .. in overall length of PJ)as I rechecked again 2day!
Plan to test it this saturday if college is off or on Sunday morning!!
Thanx once again, u really did put a serious effort with my experimental engine!! ... atleast now I know it has chances to run ... will bribe a pal of mine and try to get a vid of it as well!!
Hoping for the best ..
Rahul.
All dimensions are almost correct (+- 1mm .. in overall length of PJ)as I rechecked again 2day!
Plan to test it this saturday if college is off or on Sunday morning!!
Thanx once again, u really did put a serious effort with my experimental engine!! ... atleast now I know it has chances to run ... will bribe a pal of mine and try to get a vid of it as well!!
Hoping for the best ..
Rahul.
[url=callto://ralph_10019][/url]
[/url]-
_Rahul_
- Posts: 129
- Joined: Sat Jul 15, 2006 2:47 am
- Antipspambot question: 0
- Location: India
- Contact:
Another serious consideration ... the diameter of intake pipe is already critical... and the copper pipe that We will be using for the fuel pipe is around 4mm outer diameter... with this thick pipe inserted in the intake pipe... the net cross-section area of intake pipe is going to be less .. How far will this affect this? Won’t it be a long distance runner breathing with a straw that's partially chocked with a thinner straw?
Other option, I thought about, was putting the fuel through the tail pipe! The copper tube is pretty long... even 10 cm longer than the PJ... it can go deep into the Combustion Chamber and spit the fuel right there where its needed .. Maybe this will also improve the intake pipe dia. to the tail pipe dia. ratio! ... But then I think... the furl pipe’s going to be really hot... maybe hot enough to let the fuel burn right in the pipe... (But there won’t be any oxygen in the pipe ... so maybe no burning in there ...) but the heat can lead to some serious expansion of the fuel...!!
Won’t the same happen at the intake pipe..? Or maybe won’t happen because the fuel pipe will have less heated length at the intake pipe?
Are there any other options?
Can drilling on one side of the PJ for the fuel pipe be a good Idea...?? But I wonder how I will fix the fuel pipe in there? It can’t be welded... for in case if it needs to be removed later...
Other option, I thought about, was putting the fuel through the tail pipe! The copper tube is pretty long... even 10 cm longer than the PJ... it can go deep into the Combustion Chamber and spit the fuel right there where its needed .. Maybe this will also improve the intake pipe dia. to the tail pipe dia. ratio! ... But then I think... the furl pipe’s going to be really hot... maybe hot enough to let the fuel burn right in the pipe... (But there won’t be any oxygen in the pipe ... so maybe no burning in there ...) but the heat can lead to some serious expansion of the fuel...!!
Won’t the same happen at the intake pipe..? Or maybe won’t happen because the fuel pipe will have less heated length at the intake pipe?
Are there any other options?
Can drilling on one side of the PJ for the fuel pipe be a good Idea...?? But I wonder how I will fix the fuel pipe in there? It can’t be welded... for in case if it needs to be removed later...
[url=callto://ralph_10019][/url]
[/url]-
larry cottrill
- Posts: 4140
- Joined: Sun Oct 05, 2003 1:17 am
- Antipspambot question: 0
- Location: Mingo, Iowa USA
- Contact:
Fuel Piping
Rahul -
Go with running your fuel stinger right into the intake - it's the coolest part of your engine, period. Check out these videos of what happens to a steel tube run into the tailpipe of a pulsejet:
http://www.youtube.com/watch?v=uqXiOnWkhAY
http://www.youtube.com/watch?v=F8vqfphZ73A
Your 4mm pipe is 1/5 of the ID of the intake, so it takes up 1/25 - a whopping FOUR PERCENT of the intake area. Meaningless to engine operation.
L Cottrill
Go with running your fuel stinger right into the intake - it's the coolest part of your engine, period. Check out these videos of what happens to a steel tube run into the tailpipe of a pulsejet:
http://www.youtube.com/watch?v=uqXiOnWkhAY
http://www.youtube.com/watch?v=F8vqfphZ73A
Your 4mm pipe is 1/5 of the ID of the intake, so it takes up 1/25 - a whopping FOUR PERCENT of the intake area. Meaningless to engine operation.
L Cottrill
-
_Rahul_
- Posts: 129
- Joined: Sat Jul 15, 2006 2:47 am
- Antipspambot question: 0
- Location: India
- Contact:
Actually I am not building it upto anyone's plan .. although, I took some help from Larry's FWE. I did my own calculations and made this one .. so I aint even sure if this will run perfectly or will it give me a hard time!
Anyways, here is the pic to show how the cone was made .. just the basic drawing .. can post the detailed one with my dimensions .. but later coz I gotta run to college now!!
anyways .. here's the pic! In case u are welding the edges .. u dont need the extra edges. Hope the pic explains itself!Edit:: Just cut out the sheet of mild steel in desired shape and bend it using a wooden hammer and a wooden or steel pipe!
Anyways, here is the pic to show how the cone was made .. just the basic drawing .. can post the detailed one with my dimensions .. but later coz I gotta run to college now!!
anyways .. here's the pic! In case u are welding the edges .. u dont need the extra edges. Hope the pic explains itself!Edit:: Just cut out the sheet of mild steel in desired shape and bend it using a wooden hammer and a wooden or steel pipe!
- Attachments
-
- sample_How to make a cone.jpg
- (56.83 KiB) Downloaded 174 times
[url=callto://ralph_10019][/url]
[/url]-
_Rahul_
- Posts: 129
- Joined: Sat Jul 15, 2006 2:47 am
- Antipspambot question: 0
- Location: India
- Contact:
I used 1mm mildsteel .. I chose mildsteel becoz I get it almost free ( all scrape from the place I work .. ) .. SS is a bit costly here .. not too costly though .. but definitely out of a student's budget .. right away! I think .. SS is always better .. rust proof .. and can take more heat than Mild Steel! The PJ I made .. started getting rusted right form the next day itself .. so I have painted it to avoid further damage .. till next test run .. probably this Sunday!!
The Mild steel I used was already heat treated ..
I have already posted the technical drawing for the cone ( scalable! .. and according to the plan u are using.. ) in ur thread as well .. u can check it out!!
Hoping for the best ..
hAPPy .. pusle-jetting!!
The Mild steel I used was already heat treated ..
I have already posted the technical drawing for the cone ( scalable! .. and according to the plan u are using.. ) in ur thread as well .. u can check it out!!
Hoping for the best ..
hAPPy .. pusle-jetting!!
[url=callto://ralph_10019][/url]
[/url]-
_Rahul_
- Posts: 129
- Joined: Sat Jul 15, 2006 2:47 am
- Antipspambot question: 0
- Location: India
- Contact:
An update !!
Wanted to update this thread by the gud news that this valveless ran a week back .. I dont have a cam plus exams are holdin me back! Will post some pics and a VDO soon! Thanx everyone .. off to another PulseJet!!
[url=callto://ralph_10019][/url]
[/url]