Compression Ignition Hybrid Testbed
Posted: Sun Nov 12, 2023 10:23 am
As mentioned earlier what was presented as the Mini Hybrid II is actually a component of a wider project that I will attempt to detail here. I would first of all like to thank hectmarr with whom this project was a collaboration and without whose input would not have happened.
Here is an animation showing the basic concept, using the internals of the final iteration of the testbed more or less as built:
This combines elements of the hybrid launcher with the spring piston airgun. Rather than an electrical or chemical ignition source in a traditional hybrid, the fuel/air mixture is compressed suddenly by an externally powered piston and ignites by the diesel effect. In this case, the power for the piston came from the cartridge element from my (fl)air gun project.
Diagram of the final test setup more or less to scale:
Here are most of the testing elements that I will detail below:
A) My trusty old Beto shock pump used to pressurize the fuel/air mixture.
B) Butane fuel source.
C) Adapter made to be able to inject the fuel through the Schrader valve on the chamber.
D) Syringe used to meter fuel.
E) First endcap and barrel nut with a 1/2"-20 thread to match the thread on the air cartridge, this turned out to be too weak with a 3/8" bore and eventually cracked.
F) 6061 Aluminum chamber.
G) Fill valve, currently a Schrader but can also be replaced with a Foster type nipple. An M4 thread attaches it to the chamber and it was epoxied permanently in place with JB Weld.
H) Delrin endcap with an electrode to enable testing as a conventional hybrid.
I) 490mm long stainless steel airsoft barrel with a 6mm bore.
J) Barrel nut with an 11/16"-20 thread.
K) Limit ring between the nut and the endcap to ensure the disks are compressed in a consistent manner.
L) 11/16"-20 endcap.
M) Piston impact buffer with a stiff spring taken from a regulator and an aluminum guide.
N) UHMWPE Piston with double o-rings.
O) Piston impact buffer based on the compression of rubber washers rather than a spring that actually squeezed down the aluminum guide as a result of the impacts.
P) Examples of aluminum burst disks shown before installation, after firing and after a failure to burst from left to right.
Q) Examples of a 0.25g biodegradable airsoft BB before firing and after an impact with a steel sheet.
R) UHMWPE Piston with single o-ring and steel piston with double o-rings, the former was much kinder to the system in terms of impact but it was also felt that a heavier piston would improve performance because its momentum would better resist recoiling at the point of ignition.
S) High pressure air cartridge with steel chamber.
T) Original air cartridge chamber in aluminum that was replaced with the steel one after repeated testing with endcaps not completely tightened caused the threads to become worn.
U) Air cartridge filling adapter
For the vast majority of tests the projectile was a 0.25 gram biodegradable airsoft BB fired from a 49cm long barrel , the chamber volumes were around 12cc for the air cartridge and 28cc for the cylinder with the spring buffer installed. As burst disk material I was using either aluminum cut from the side of a beer can or photo paper. When pumped until failure, a single layer of the former burst at around 1500 psi. Other variables include piston weight, pre-pressurization levels in the second chamber, pressure in the air cartridge and different fuel types to achieve dieseling. Here is the data obtained:
There were many different variables tested so there is quite a lot to say about it, I will try and condense it in some salient points.
- As a baseline, I connected the barrel directly to the air cartridge to evaluate the performance if the second chamber was completely eliminated. The results can be seen on lines 12 and 13, 18.8 and 21.6 ft lbs for 1500 psi and 2000 psi respectively. Pumping the second chamber with a single aluminum disk until it burst yielded 17.2 ft lbs on line 11, it's interesting that in spite of the fact that the air cartridge chamber is less than half the size of the second chamber and the burst disk is supposed to be the ultimate valve with maximum flow and zero opening time from the outset, there was no significant difference in energy.
- On line 2 using a 5.5 gram Delrin/aluminum piston and 1500 psi in the air cartridge with no fuel in or pre-pressurization of the second chamber I got a mere 2.9 ft lbs. The aluminum disk burst, that means pressures reached at least the 1500 psi ballpark under compression, but from a much tinier volume. It's worth noting that there is a not inconsiderable amount of dead volume caused by the spring buffer and this limits the maximum amount of pressure that can be achieved. Using the same parameters but adding different fuels in the second chamber, a significant boost in velocity was obtained, showing the difference fuel ignition makes beyond simply compressing the air.
- On lines 8, 9 and 10 I switched to a 3.1 gram UHMWPE piston, and increasing the air cartridge pressure and therefore the speed at which the air in the second chamber is compressed yielded a progressive increase in energy but not a dramatically different one.
- On line 14, adding just 30 psi to the second chamber (effectively tripling the amount of air) yielded a dramatic increase in energy, 8.1 ft lbs compared to 2.9 ft lbs. Subsequent tests with gas fuel added on lines 15 and 16 did not yield a similar increase, but I have reason to believe that due to leaks the concentration of gas was not within the ignition limits and it did not burn. With a drop of cooking oil on line 16 however, adding 45 psi of pre-pressurization boosted the energy to 12.0 ft lbs up from 6.2 ft lbs, just under double.
- On line 18 I switched to a new 3.8 gram UHMWPE piston with double o-rings and sorted out some small leaks in the pump as well as adding a ring between the barrel nut and the endcap to ensure that the burst disk was always torqued to the same level. This is important because if it is not enough, the disk can slip rather than burst, and if it is too tight, then the o-ring compression will expose a lower area of the disk to the pressure and it will take more to burst it. As a result of these changes I am more confident in the results after this point involving pre-compression in the second chamber. With 60 psi of pre-pressurization the energy went up to 8.8 ft lbs with butane from 4.9 ft lbs with air only.
- On lines 20 and 21 I repeated the previous tests but using two layers of burst disk instead of one. The energy without the fuel was higher, but interestingly the energy with the fuel was the lowest of all four tests.
- On line 22 increasing pre-pressurization by 50% to 90 psi with fuel yielded 10.0 ft lbs up about 56% from the test at 60 psi.
- On line 23 I used the second chamber alone, using the endcap H) and filling it to 205 psi with 5% concentration Butane and ignited it using a grill sparker. This yielded a velocity just under twice the speed of sound and 41.8 ft lbs of energy, far beyond what the piston compression setup was achieving. This is hardly surprising because the volume in the chamber after compression is very small while in this case the entirety of the chamber was pressurized to a relatively high level, with 15 times more butane than the tests at atmospheric pressure.
- On line 24 I repeated the test on line 19 but using a steel piston at 24.7 grams, almost 8 times heavier. This yielded an increase in energy at 12.4 ft lbs up from 8.8 ft lbs, and I suspect this is due to the fact that the piston might accelerate slower but also decelerates less when the mixture ignites, meaning a higher sustained pressure in the system during firing.
- On lines 25 and 26, I repeated lines 18 and 19 but using photo paper as a burst disk, a weaker material that would burst at lower pressure. Interestingly this yielded a higher velocity without fuel that the aluminum disk with fuel, but also almost no difference between the result with and without fuel. I suspect the disk burst before enough pressure was reached for the fuel to ignite.
- On lines 27 and 28 I tested the photo paper disk with electric only ignition, it did not burst with a 5x atmospheric pressure mix but it did at 10x.
- On lines 29 to 33 I did some more testing with the steel piston, the best result was with cooking oil at 17.2 ft lbs and 130 psi pre-pressurization of the second chamber, up from 15.1 at 60 psi. Increasing this to 215 psi on line 33 yielded less than half the energy, likely because of the increased resistance to piston acceleration. Using two burst disks on line 32 and a lighter piston on line 34 yielded significantly less energy.
- On lines 35 and 36 I tried with electric ignition at 17x and 19x atmospheric pressure and got velocities well in excess of Mach 2 with 0.25g airsoft BBs. The 19x test was repeated on line 37 using a 1.67 gram 0.25" lead pellet swaged down to 6mm and it yielded a whopping 88.6 ft lbs, just over 50% more energy at the muzzle compared to the BB using the same parameters.
- On lines 38 to 41 I again used electric ignition to see how different burst disk layers would affect performance. Using two disks yielded almost 75% more energy than one disk. Three disks also yielded more but only around 40%, and with four disks they did not burst at all.
All in all it was fun series of tests fiddling with variables that I hope will be interesting and slinging plastic at over twice the speed of sound is also a rather gratifying experience
Here is an animation showing the basic concept, using the internals of the final iteration of the testbed more or less as built:
This combines elements of the hybrid launcher with the spring piston airgun. Rather than an electrical or chemical ignition source in a traditional hybrid, the fuel/air mixture is compressed suddenly by an externally powered piston and ignites by the diesel effect. In this case, the power for the piston came from the cartridge element from my (fl)air gun project.
Diagram of the final test setup more or less to scale:
B) Butane fuel source.
C) Adapter made to be able to inject the fuel through the Schrader valve on the chamber.
D) Syringe used to meter fuel.
E) First endcap and barrel nut with a 1/2"-20 thread to match the thread on the air cartridge, this turned out to be too weak with a 3/8" bore and eventually cracked.
F) 6061 Aluminum chamber.
G) Fill valve, currently a Schrader but can also be replaced with a Foster type nipple. An M4 thread attaches it to the chamber and it was epoxied permanently in place with JB Weld.
H) Delrin endcap with an electrode to enable testing as a conventional hybrid.
I) 490mm long stainless steel airsoft barrel with a 6mm bore.
J) Barrel nut with an 11/16"-20 thread.
K) Limit ring between the nut and the endcap to ensure the disks are compressed in a consistent manner.
L) 11/16"-20 endcap.
M) Piston impact buffer with a stiff spring taken from a regulator and an aluminum guide.
N) UHMWPE Piston with double o-rings.
O) Piston impact buffer based on the compression of rubber washers rather than a spring that actually squeezed down the aluminum guide as a result of the impacts.
P) Examples of aluminum burst disks shown before installation, after firing and after a failure to burst from left to right.
Q) Examples of a 0.25g biodegradable airsoft BB before firing and after an impact with a steel sheet.
R) UHMWPE Piston with single o-ring and steel piston with double o-rings, the former was much kinder to the system in terms of impact but it was also felt that a heavier piston would improve performance because its momentum would better resist recoiling at the point of ignition.
S) High pressure air cartridge with steel chamber.
T) Original air cartridge chamber in aluminum that was replaced with the steel one after repeated testing with endcaps not completely tightened caused the threads to become worn.
U) Air cartridge filling adapter
For the vast majority of tests the projectile was a 0.25 gram biodegradable airsoft BB fired from a 49cm long barrel , the chamber volumes were around 12cc for the air cartridge and 28cc for the cylinder with the spring buffer installed. As burst disk material I was using either aluminum cut from the side of a beer can or photo paper. When pumped until failure, a single layer of the former burst at around 1500 psi. Other variables include piston weight, pre-pressurization levels in the second chamber, pressure in the air cartridge and different fuel types to achieve dieseling. Here is the data obtained:
- As a baseline, I connected the barrel directly to the air cartridge to evaluate the performance if the second chamber was completely eliminated. The results can be seen on lines 12 and 13, 18.8 and 21.6 ft lbs for 1500 psi and 2000 psi respectively. Pumping the second chamber with a single aluminum disk until it burst yielded 17.2 ft lbs on line 11, it's interesting that in spite of the fact that the air cartridge chamber is less than half the size of the second chamber and the burst disk is supposed to be the ultimate valve with maximum flow and zero opening time from the outset, there was no significant difference in energy.
- On line 2 using a 5.5 gram Delrin/aluminum piston and 1500 psi in the air cartridge with no fuel in or pre-pressurization of the second chamber I got a mere 2.9 ft lbs. The aluminum disk burst, that means pressures reached at least the 1500 psi ballpark under compression, but from a much tinier volume. It's worth noting that there is a not inconsiderable amount of dead volume caused by the spring buffer and this limits the maximum amount of pressure that can be achieved. Using the same parameters but adding different fuels in the second chamber, a significant boost in velocity was obtained, showing the difference fuel ignition makes beyond simply compressing the air.
- On lines 8, 9 and 10 I switched to a 3.1 gram UHMWPE piston, and increasing the air cartridge pressure and therefore the speed at which the air in the second chamber is compressed yielded a progressive increase in energy but not a dramatically different one.
- On line 14, adding just 30 psi to the second chamber (effectively tripling the amount of air) yielded a dramatic increase in energy, 8.1 ft lbs compared to 2.9 ft lbs. Subsequent tests with gas fuel added on lines 15 and 16 did not yield a similar increase, but I have reason to believe that due to leaks the concentration of gas was not within the ignition limits and it did not burn. With a drop of cooking oil on line 16 however, adding 45 psi of pre-pressurization boosted the energy to 12.0 ft lbs up from 6.2 ft lbs, just under double.
- On line 18 I switched to a new 3.8 gram UHMWPE piston with double o-rings and sorted out some small leaks in the pump as well as adding a ring between the barrel nut and the endcap to ensure that the burst disk was always torqued to the same level. This is important because if it is not enough, the disk can slip rather than burst, and if it is too tight, then the o-ring compression will expose a lower area of the disk to the pressure and it will take more to burst it. As a result of these changes I am more confident in the results after this point involving pre-compression in the second chamber. With 60 psi of pre-pressurization the energy went up to 8.8 ft lbs with butane from 4.9 ft lbs with air only.
- On lines 20 and 21 I repeated the previous tests but using two layers of burst disk instead of one. The energy without the fuel was higher, but interestingly the energy with the fuel was the lowest of all four tests.
- On line 22 increasing pre-pressurization by 50% to 90 psi with fuel yielded 10.0 ft lbs up about 56% from the test at 60 psi.
- On line 23 I used the second chamber alone, using the endcap H) and filling it to 205 psi with 5% concentration Butane and ignited it using a grill sparker. This yielded a velocity just under twice the speed of sound and 41.8 ft lbs of energy, far beyond what the piston compression setup was achieving. This is hardly surprising because the volume in the chamber after compression is very small while in this case the entirety of the chamber was pressurized to a relatively high level, with 15 times more butane than the tests at atmospheric pressure.
- On line 24 I repeated the test on line 19 but using a steel piston at 24.7 grams, almost 8 times heavier. This yielded an increase in energy at 12.4 ft lbs up from 8.8 ft lbs, and I suspect this is due to the fact that the piston might accelerate slower but also decelerates less when the mixture ignites, meaning a higher sustained pressure in the system during firing.
- On lines 25 and 26, I repeated lines 18 and 19 but using photo paper as a burst disk, a weaker material that would burst at lower pressure. Interestingly this yielded a higher velocity without fuel that the aluminum disk with fuel, but also almost no difference between the result with and without fuel. I suspect the disk burst before enough pressure was reached for the fuel to ignite.
- On lines 27 and 28 I tested the photo paper disk with electric only ignition, it did not burst with a 5x atmospheric pressure mix but it did at 10x.
- On lines 29 to 33 I did some more testing with the steel piston, the best result was with cooking oil at 17.2 ft lbs and 130 psi pre-pressurization of the second chamber, up from 15.1 at 60 psi. Increasing this to 215 psi on line 33 yielded less than half the energy, likely because of the increased resistance to piston acceleration. Using two burst disks on line 32 and a lighter piston on line 34 yielded significantly less energy.
- On lines 35 and 36 I tried with electric ignition at 17x and 19x atmospheric pressure and got velocities well in excess of Mach 2 with 0.25g airsoft BBs. The 19x test was repeated on line 37 using a 1.67 gram 0.25" lead pellet swaged down to 6mm and it yielded a whopping 88.6 ft lbs, just over 50% more energy at the muzzle compared to the BB using the same parameters.
- On lines 38 to 41 I again used electric ignition to see how different burst disk layers would affect performance. Using two disks yielded almost 75% more energy than one disk. Three disks also yielded more but only around 40%, and with four disks they did not burst at all.
All in all it was fun series of tests fiddling with variables that I hope will be interesting and slinging plastic at over twice the speed of sound is also a rather gratifying experience
