Blast Compression Intake proposed design

Moderator: Mike Everman

Ogge
Posts: 109
Joined: Fri Jan 07, 2005 8:18 pm
Antipspambot question: 0
Location: Alpharetta, GA

Blast Compression Intake proposed design

Post by Ogge » Mon Jan 24, 2005 11:13 am

I am moving part of my latest post in the CC design thread to a new thread for seperate discussion. This is my proposed design for generating blast compression and transverse waves to compress the CC and initiate ignition timing under designer control limiting the requirements for precise resonance timing.

I am designating the term BCI to represent my "Blast Compression Intake" assembly. They are designed to be used in a thermojet configuration PDE or pulsejet. See my "Proposed CC Design" thread.

The BCI is designed to be used in the place of a normal intake tube. It consists of a lenght of straight small diameter tube that runs parallel to the exhaust tube with a T intersection that is the actual separate intake tube of very short lenght with a flared end.

Multiple BCI tubes should be angled into the CC to converge the blast fronts to intersect at the chosen ignition point in the CC. Since it angles into the CC, it needs to be bent so it can run parallel to the exhaust tube. This bend is actually beneficial, it will induce rotation (like the currents in a meandering river) in the exhaust gasses providing a small tornado of hot gases into the CC. This will cause rapid widespread combustion in the CC.

BCI tubes should be a little longer than the exhaust with a conical end plug to compress, ignite a fuel mixture there. The timing for the reflective shockwave should arrive at about the correct time for reigntion of follow on cycle. The "actual" intake will have to be split off from these Blast compression tubes ~1/4 of the way from the CC to the end plug. This short intake tube will be closest to the CC end of the BCI.

I envission the intakes coming in at 90 degrees off this tube and curved to run parallel to it. The curve is required to induce vortex rotation to draw mixture in faster. It will be required because the intake must navigate a 90 degree turn into the BC tube when recharging the dead end branch or refreshing the CC. This intake tube needs to be short. It will need a high resonance frequency. The waves from its oscilation should not really propogate back into the CC due to the 90 degree intersection. This oscilation will operate like a high speed pump.

Theory of Operation:
The CC ignites, drives shockwave down both the exhaust and the BCI tubes. The wave in the BCI tube will pass the intake port at ~1/4L position. This will vent some of the back pressure but the initial shockwave will not make the 90 degree turn and momentum of the exhaust gasses will carry alot of the expasion gassed straight down to the dead end of the BCI tube.

The shockwave will compress the mixture in the BCI tube dead end, reflect off the conical end-cap (focusing the wave) and drive the mixture directly into the flamewave that followed the shockwave. Whether this mixture ignites by the shockwave focusing alone or ignites when the flamewave intersects is basicly irrelevant. This charge has only 1 exit path, back towards the CC.

By the time the enhanced reflected blast is starting to return, the CC will be at its low pressure point sucking in mixture thru the intake. This is due to the fact that the BCI is longer then the exhaust tube. The very short intake tube branched from the BCI will be resonating at a high frequency pumping multiple layers of mixture into the top (CC end of the BCI) and being drawn into the CC. The top 1/4 section of the BCI should also be close to vacuum due to overexpassion by gas momentum.

One reason Im using 3 to 5 intakes thermojet style is to allow enough intake volume during this reduced timeframe for recharging the CC. The other reason is to drive the intake mixture into the closed dead end of the CC frontwall.

The reflected blast from the BCI driven by the small secondary combustion in the dead end of the BCI will drive pass the intake port blocking it. It will vent some of its back pressure but the majority of the flow should be driving into the CC due to momentum and the lower pressure in the CC.

At this same time, the reflectioned waves from the main exhaust tube will be drawing in the fresh cold air mass from the exhaust driving up the CC pressure. The BCI shockwave will DRIVE the remaining mixture still in the top 1/4 of the tube into the CC compressing it into the nose-cone with a flamewave following right behind it.

Once the BCI overexpands, a fresh mixture is drawn from the intake refreshing the BCI dead end section. This is occuring while 3-5 transverse blast waves intersect into the CC at the ignition point into a mass of highly compressed layers of mixture with a flame front driving right behind it thus reigniting the CC.

Resonance timing from the exhaust reflections do not have to be timed to intersect perfectly. There should be no need for exact timing. The BCI blast front should be plenty of energy to reignite the CC. As long as the cold air mass has been reloaded into the exhaust tube. That and helping to overpressurize the CC is all that is needed from the exhaust reflections.

This set up should provide much higher CC compression. It should provide layers of mixture drawn into the CC all compressed into the dead end nose cone of the CC.

The only problems I forsee are being able to draw in enough mixture during the shortened refresh cycle thus the need for multiple intakes and the vortex induced by curving the short intake tube. This has the added benefit of providing transverse intersecting wavefronts. Another forseeable problem would be from timing but the ignition timing can be varied by moving the BCI conical end plug closer towards the CC. The final problem will be preventing the entire thing from reaching DDT by the high compression in the CC or BCI dead ends when used in a pulsejet of limited tube thinkness.

I got my inspiration from Bruno's BCVP (Thanks Bruno). But I see several problems with his proposal. The intake and exhaust have to share the same tube. The BC is not driving into a dead end, therefore it is much harder to get good compression. His design is also just a long narrow CC that doesnt have time to refresh. By placing the intake between the blast from a large CC and a small dead end reflection and splitting the BC into multiple smaller chambers as well as providing a clear separate path for exhaust has helped to solve those issues.

One other thing I would like to point out. If the BCIs dont allow the CC to breath enough, you could also use some normal Intake tubes. For example: a 5 intake port thermojet with 2 normal intake tubes and 3 BCI tubes configured in a star configuration or 6 ports of 3 and 3 in a hexagonal configuration.
Last edited by Ogge on Tue Jan 25, 2005 1:02 pm, edited 1 time in total.
Adam Becker
Innovative Propulsion

Dave
Posts: 227
Joined: Sat Nov 08, 2003 7:03 pm
Antipspambot question: 0
Location: Buffalo, NY

Re: Blast Compression Intake proposed design

Post by Dave » Mon Jan 24, 2005 11:43 am

Ogge
I think I get most of what you are proposing, but am still having some difficulty visualizing the placement of the 90 degree turns. If I understand the theory correctly, you are proposing a variation on what Bruce Simpson appears to have done with his mystical X-Jet engine. It would be very helpful if you could provide a sketch to go along with your description.
Dave

Ogge
Posts: 109
Joined: Fri Jan 07, 2005 8:18 pm
Antipspambot question: 0
Location: Alpharetta, GA

Re: Blast Compression Intake proposed design

Post by Ogge » Mon Jan 24, 2005 12:24 pm

Dave wrote:Ogge
I think I get most of what you are proposing, but am still having some difficulty visualizing the placement of the 90 degree turns. If I understand the theory correctly, you are proposing a variation on what Bruce Simpson appears to have done with his mystical X-Jet engine. It would be very helpful if you could provide a sketch to go along with your description.
Dave

[Edited]

See post of picture in later post.
Last edited by Ogge on Thu Jan 27, 2005 7:08 am, edited 1 time in total.
Adam Becker
Innovative Propulsion

larry cottrill
Posts: 4140
Joined: Sun Oct 05, 2003 1:17 am
Antipspambot question: 0
Location: Mingo, Iowa USA
Contact:

Re: Blast Compression Intake proposed design

Post by larry cottrill » Mon Jan 24, 2005 1:56 pm

Adam -

A way, way back on the old forums, my very first valveless pulsejet design, the Synchrodyne(TM) attempted to use a compression chamber, but I never achieved any ignition there, and the engine did not sustain. Of course, if I were doing it today, I think the proportions of the volumes and lengths would be a lot different.

Though something of an embarrassment, and definitely an 'immature work' [i.e. I didn't know squat], it can still be seen at:
http://www.cottrillcyclodyne.com/syncrdy1.html
with a log of part of my testing at:
http://www.cottrillcyclodyne.com/synctst1.html

The basic idea is shown below, as it was after I added a non-resonant [short] forward-facing intake - also shown is a redesign that might have produced better results [or, maybe not].

L Cottrill
Attachments
Synchrodyne_II.jpg
Redesign "in the round", using CO2 cylinders, my favorite construction material. Drawing Copyright 2003 Larry Cottrill
(53.74 KiB) Downloaded 1158 times
synchrodyne 4.jpg
Not-to-scale drawing of the original Synchrodyne(TM) design, after addition of forward-facing intake. This is basically the final tested configuration of the prototype. Drawing Copyright 2001 Larry Cottrill
synchrodyne 4.jpg (83.61 KiB) Viewed 20312 times

Ogge
Posts: 109
Joined: Fri Jan 07, 2005 8:18 pm
Antipspambot question: 0
Location: Alpharetta, GA

Re: Blast Compression Intake proposed design

Post by Ogge » Mon Jan 24, 2005 3:02 pm

Larry Cottrill wrote:Adam -

A way, way back on the old forums, my very first valveless pulsejet design, the Synchrodyne(TM) attempted to use a compression chamber, but I never achieved any ignition there, and the engine did not sustain. Of course, if I were doing it today, I think the proportions of the volumes and lengths would be a lot different.
L Cottrill
Larry,
Im not surprized that neither of them would work. Remember what I have said about wave focusing. When you ignite inside a cone or even a partial cone, the net vector force is from the narrow end of the cone directly down the axis towards the wide end. A lockwood CC uses a partial conical section on either end of the CC attempting to focus part of the energy down both tubes.

Your FWE design with the long gradual cone is focusing most of the waves into the front CC cap. I would bet money that if you reversed the conical sections on the FWE CC you would get an increase in thrust. The longer the cone the better for focusing. This can be proved easily mathmaticly or you can research shape charges.

For example, in drilling for oil, it is sometimes necessary to blast open the oil bearing rocks deep in the well. In the old days they use to just lower dynamite into the hole but the explosion was not directed, it could damage or block the well. A college friend of my father wrote a research paper on the properties of cones to direct force for application in oil wells. His professor stole the idea and patented it and got rich. His designs are commonly still used today by the oil industry.

Another example is military ordenance for tankbusters, they pack a block of steal with a conical section bored out. Place high explosives in the cone and cap it with a metal disc. The force of the explosion turns the cap disc into a sharply pointed projectile for armor piercing.

http://www.systems.textron.com/movies/sfw_a.wmv

Here is the simulation of a partial cone shape charge. Check out what happens to the dark red area. It is there to show the resulting force vector. Note how the point starts out towards the back but gets extended greatly down the center axis. In a CC you have a confined combustion just like a shapecharge.

http://www.feainformation.com/avilib/62.avi

Another example is the
http://www.aerospaceweb.org/design/aero ... apes.shtml
Here is the quote where they talk about conical nozzles for rockets: "A small angle produces greater thrust, because it maximizes the axial component of exit velocity and produces a high specific impulse".

Both of your designs are attempting to focus away from your compression chambers. You will get very little force in that direction. Only the so called "constant pressure expansion". Basicly it would just work like a shock obsorber for the expansion providing no useful benefit, they would detract from performance.

My design is using focused shockwaves driving a into a cone, compressing the fuel/mixture smaller and smaller into the tip of the cone. Combustion will progress clear into the dead-end BCI because the entire tube should be filled with mixture and initiate a very energetic reaction. Im not at all concerned that my design will not combust.

Hopefully it will not Detonate! But this is one way to trigger full DDT.

[Edited and appended]

Larry,
after re-reading my post, I felt that it sounded a bit to derogatory. That was not my intention.

I applaud your thinking to try new things and to have already attempted to achieve the same type of thing I am attempting. It just that by looking at your designs I see so many things that go against my understanding of pulsejet operations. I feel compelled to point them out. My spiritual gift is teaching, so please dont let it bother you when I point out flaws as I see them.

So after looking at your designs, please note that you are dividing your expansion in 2 directions. Only half can go either way, then you are focusing a large portion of the force that would have gone to the reflection chamber to the exhaust with your narrowing conical section. Unlike your design, mine is focusing all energy in 1 direction, allowing it to accelerate more rapidly then it taps off a small portion of it to use in a much more confined reflection chamber that has no impedance to flow, allows a reserve of unconsumed mixture of fuel to be compressed by the leading edge of the pressure wave and concentrates it.

I see very little in common with your designs except for that fact (I am assuming) that you had similar conseptual idea and it didnt work.
Last edited by Ogge on Wed Jan 26, 2005 10:25 pm, edited 3 times in total.
Adam Becker
Innovative Propulsion

mk
Posts: 1053
Joined: Mon Jan 19, 2004 8:38 pm
Antipspambot question: 0
Location: FRG

Re: Blast Compression Intake proposed design

Post by mk » Mon Jan 24, 2005 8:35 pm

Ogge wrote:Here is the quote where they talk about conical nozzles for rockets: "A small angle produces greater thrust, because it maximizes the axial component of exit velocity and produces a high specific impulse".
Hey, I got the point! The importance of wave-action was really so obscure to me all the time...thats why augmenters/ejectors should not exceed a 4deg angle...cool.

Adam, you are, however, considering wave-action first and caring about mass flow more or less later on, right?
Just the opposed way of thinking to the main part of the people here (still including me).
Maybe the wave-action is the key for Forrest's ("WebPilot") computer modeling work of pulsejets?

You can really learn so many new things here every day.
mk

Ogge
Posts: 109
Joined: Fri Jan 07, 2005 8:18 pm
Antipspambot question: 0
Location: Alpharetta, GA

Re: Blast Compression Intake proposed design

Post by Ogge » Mon Jan 24, 2005 10:20 pm

mk wrote: Adam, you are, however, considering wave-action first and caring about mass flow more or less later on, right?
Marten,
That really depends on your definition of mass flow. If your talking about the mass flow from the so-called "constant pressure expansion" or "actual" mass flow due to all factors.

As I see it, the wave-action is the root cause of the "actual" mass flow, what initiated it, what accelerates it. Its the driving force of the tempature and the pressure which creates the mass flow.

So in that respect, I would have to say no, I am VERY concerned about mass flow. I am just doing everything in my power to maximize mass flow by maximizing concentrations of energy waves to drive mass flow in the direction of my chosing, that is the ultimate goal.

Here is an example: if you just let random waves bounce arround a poorly designed CC, you are simply raising the tempature in the volume of the CC and generating hot spots in the walls. That tempature pushs somewhat evenly at all areas that expansion are possible thus the term "constant pressure expansion".

What I am doing is focusing the waves to drive right into that initial expansion to add energy. That energy in heat/molecular or electromagnetic vibrations has an effect on the boundary condition. It generates more heat/pressure concentrated at that boundary thus moving it faster, transfering more heat to the cold mass, increasing the amplitude of the leading edge. But all of this is done only to the boundary area I want to do this too. Mainly the exhaust port boundary.

"Constant pressure expansion" is a myth as far as Im concerned. It doesnt really apply to pulsejets IMHO. "Constant pressure expansion" is an average of an area of confined boundaries. This average is useful in some cases. Examples of this figure being useful are in devices with basicly constant pressure/tempature outputs like continous combustion devices (turbo-jets, rockets, boilers, ect). Due to the continous combustion of those devices, the average is the most important factor.

In pulsejets, there is never a constant condition. We never reach a steady state. Waves are crisscrossing all the chambers/tubes of the pulsejet in all directions altering the pressure/tempature gradients at any given point constantly. In fact we count on them being there to continue the cycle.

Everyone knows we want standing waves, but few remember to have a standing wave you must have multiple continous waves traveling in opposite directions at equal magnitudes to generate these standing waves. Everyone is focusing on the sum of all the waves as opposed to the individual waves that make up that sum.

To get a true standing wave you must have exact multples of a given frequency traveling perfectly down a constant condition environment. Basicly this can only be done in a constant tempature environment in a single open ended tube or a tube closed on one end such as a big cathedral pipe organ. We dont generate a single constant tone out of a pulsejet. We cant. We have brawd ranging noise over a wide spectrum of frequencies. What has been attempted is to try to make certain frequencies dominate in a chaotic environment.

We will never truely be close to having clean standing waves. We may get a pressure spike at a key time at a key location that can trigger reignition but thats about as good as its gonna get with current designs.

Now with regard to this design. I am trading some of the energy that would have gone down the exhaust to do other work for me instead. What I am doing is to directly that energy to override the concept of attempting to use the reflection from the exhaust as the key driving factor that limits our performance.

I am replacing it would something I have more direct control over, that can be easily tuned to occure when I want it to occur. By doing this, I can truely maximize the ignition sequence getting greater efficiency overall. I can drive it faster, or slower, I can get the energy to release in more concentrated form so I can focus it more directly. I am trading barely understood chaos for order.

With this order, the engine will have a much broader range of operation, with larger variations to fuel mixtures, or fuel types. It will have much greater stability. It will not reach a tempature that will drive the so called standing waves so far out of sinc that the engine just stops.

I am doing this by replacing the need to rely on the exhaust reflection speed to arrive at a precise time to continue operation. I generate a more powerful reflective wave to drive the ignition. With this being done, the exhaust reflections role become much less critical. Its only true role is to ensure that a new cold air mass is somewhat loaded into the exhaust tube.

As a matter of fact, I can eliminate the reflection wave entirely turning that energy into thrust instead and let pure suction pull in the cold air mass or vent the exhaust tube closer to the CC to allow a shorter recharge path. The exhaust reflection can be prevented by generating a more matched impendance at the exhaust. Its the impedance mismatch that caused the reflection. To match impedance, you need to match pressure and limit the tempature gradient. You can do this with nozzle designs by expanding the exhaust mass with a gradual increase in exhaust port size like a lockwood or with a bell shaped exhaust port like on the space shuttle.

By setting multiple BCI tubes to work in parallel, I am ensuring that the all the required waves for reignition are traveling at the same speeds, in the same environment under all conditions. The exhaust or intake tubes can be given lengths that are more user friendly, make them shorter if I wish, or as long as I want (within reason). The BCI tubes can be tuned easily to generate a larger range of frequency output limited only by thermal conditions.

And last but not least, I am designing an engine that can be a pulsejet or a PDE. The only thing needed to turn this into a PDE would be a strong enough pulse to get sustanable detonation in the CC to the exhaust and propogate enough force into the BCI tube to reflect back another trigger force strong enough to repeat the cycle. Of course there are other limiting factors to sustaining a PDE that is based on physical tube diameters, chose of fuels, ect...

I entend to build an engine similar to what I have preposed, fuel it with specific fuels and inject a high explosive trigger charge and generate PDE mode. Of course I will have to make the engine much more powerful to withstand those forces. I will basicly mill the entire engine out of solid slabs of high tempature alloys and bolt it together very well. Instead of having an exhaust tube, it will be a hole bored into center an 6-8" solid peice of bar stock with the other tubes bored into it farther from the center.

Many of you may disagree with my methods or the statement I have just made. Feel free to tear it appart and point out any flaws as you see them. Im sure I made many wrong assumptions and will have to rethink a few things.
Adam Becker
Innovative Propulsion

mk
Posts: 1053
Joined: Mon Jan 19, 2004 8:38 pm
Antipspambot question: 0
Location: FRG

Re: Blast Compression Intake proposed design

Post by mk » Tue Jan 25, 2005 6:52 pm

Ogge wrote:Everyone knows we want standing waves, but few remember to have a standing wave you must have multiple continous waves traveling in opposite directions at equal magnitudes to generate these standing waves. Everyone is focusing on the sum of all the waves as opposed to the individual waves that make up that sum.
When a (nearly) perfect "normal" pulsejet is a xenon lamp, than your aim is building an engine being the laser under pulse combustion devices, eh?
(The generation of a standing wave through a lot of "small" waves reminds me to the operation of a laser.)

Seeing the day to come this year, which I plan to start studying: Are these things more theoretical physics stuff or more aviation/space_travel issues?
(I'm still alternating between two main subjects.)
mk

Ogge
Posts: 109
Joined: Fri Jan 07, 2005 8:18 pm
Antipspambot question: 0
Location: Alpharetta, GA

Re: Blast Compression Intake proposed design

Post by Ogge » Tue Jan 25, 2005 7:46 pm

Anyone have a way to convert a varicad drawing to jpg so I can post the design? My version of varicad has expired and I cant export it to another drawing type.
Adam Becker
Innovative Propulsion

Ogge
Posts: 109
Joined: Fri Jan 07, 2005 8:18 pm
Antipspambot question: 0
Location: Alpharetta, GA

Re: Blast Compression Intake proposed design

Post by Ogge » Wed Jan 26, 2005 4:27 am

mk wrote: Seeing the day to come this year, which I plan to start studying: Are these things more theoretical physics stuff or more aviation/space_travel issues?
(I'm still alternating between two main subjects.)
Actually its Electrical Engineering. What other field deals with waves (A/C electricity,Radio/TV,fiber optics) more? Most of our training is in wave theory, manipulation and oscilation and how to get it to do what you want. All repeating waveform are made up of a series of sine waves. For example, a squarewave (0101010101) is actually made up of a primary frequency sine wave + odd harmonics (multiples of that primary frequency). We even have to deal with sine waves working in digital electronics.

A pulsejet is just a simple oscilator, granted it is oscilating at alot of different frequencies that vary over time. This is similar to a spread spectrum transmiter. It is also like what you recieve at your FM antenna, broad spectrum noise. In an FM reciever, you build a device to lock on to the frequency you want, filter out all others and extract the information in the signal, amplify it and send it to your speakers.

Basicly, I am treating a pulsejet as an unreliable oscillator. If you have an unreliable oscillator in the electrical realm, you steady it with a phase lock loop (basicly how an FM reciever works) or you use a stable outside oscillator to drive the circuit. Thats what I am attempting to do.

If the oscilation in a pulsejet is broken or unrealiable, quit trying to fix it, just filter out the noise and provide a stable oscilation.
Adam Becker
Innovative Propulsion

Bruno Ogorelec
Posts: 3542
Joined: Sat Oct 04, 2003 7:31 am
Antipspambot question: 0
Location: Zagreb, Croatia

Re: Blast Compression Intake proposed design

Post by Bruno Ogorelec » Wed Jan 26, 2005 9:33 am

Ogge wrote:Basicly, I am treating a pulsejet as an unreliable oscillator. If you have an unreliable oscillator in the electrical realm, you steady it with a phase lock loop (basicly how an FM reciever works) or you use a stable outside oscillator to drive the circuit. Thats what I am attempting to do.

If the oscilation in a pulsejet is broken or unrealiable, quit trying to fix it, just filter out the noise and provide a stable oscilation.
Ahem... I seem to remember attempts by whole groups of (rather smart) forum participants to convert pulsejet behavior to electrical analogues, in order to be able to handle the relations more easily and get a clearer picture. I may be wrong, but I don't remember them getting anywhere.

Stream of electrons is a stream of electrons. Stream of molecules of fluid is a stream of molecules of fluid. Similarities only go part of the way -- the easy part. It is where the two are different that the problems start.

I've said that before to other people trying to switch to electrical equivalents. Don't do it. It's a false path strewn with traps.

I've never understood the effort to study birds and bees in order to understand sex. There are more direct ways. :o)

Dave
Posts: 227
Joined: Sat Nov 08, 2003 7:03 pm
Antipspambot question: 0
Location: Buffalo, NY

Re: Blast Compression Intake proposed design

Post by Dave » Wed Jan 26, 2005 11:59 am

Guys
There is a way to settle all this back and forth theoretical discussion.
Build, test and let the results speak for themselves.
Dave

steve
Posts: 1029
Joined: Sun Jun 06, 2004 12:29 am
Antipspambot question: 0
Location: Clinton Conneticut / Melbourne Flordia
Contact:

Re: Blast Compression Intake proposed design

Post by steve » Wed Jan 26, 2005 2:36 pm

Ogge wrote:Anyone have a way to convert a varicad drawing to jpg so I can post the design? My version of varicad has expired and I cant export it to another drawing type.
open up the program so that you have a good view of what ever you want to show us, then use the "Print Screen" button on your keyboard. This will take a screen shot and copy it to the clipboard. Then simply open up Paint or any other immage editing software and paste ithe immage. Then you can crop it etc. and save it as a jpg.

This is the method I use for every CAD drawing I've ever posted here, so it works quite well.
Image

Ogge
Posts: 109
Joined: Fri Jan 07, 2005 8:18 pm
Antipspambot question: 0
Location: Alpharetta, GA

Re: Blast Compression Intake proposed design

Post by Ogge » Wed Jan 26, 2005 3:33 pm

Thanks Steve, that worked very nicely.
Last edited by Ogge on Sat Jan 29, 2005 12:26 am, edited 2 times in total.
Adam Becker
Innovative Propulsion

Ogge
Posts: 109
Joined: Fri Jan 07, 2005 8:18 pm
Antipspambot question: 0
Location: Alpharetta, GA

Re: Blast Compression Intake proposed design

Post by Ogge » Wed Jan 26, 2005 4:06 pm

deleted
Last edited by Ogge on Sat Jan 29, 2005 12:26 am, edited 1 time in total.
Adam Becker
Innovative Propulsion

Post Reply