English Electric Thunderbird build

[delta22]

patelldp,

There are many relevant issues your previous comments fail to consider:

1) For relatively slow rockets, the relationship between altitude and impulse is very far from linear. For example my son's LOC Viper 4 flies 650ft on 4x D12 and 320 ft on 3x D12. This is with 7 oz rocket weight per D12, for the Thunderbird at 10 oz per D12, the proportional reduction in altitude is greater.

2) Estes motors have a substantial thrust spike at the beginning of the burn that I counted on to successfully to achieve a straight vertical flight off the 6 ft rail.

3) There were very light winds on the field at the time of launch.

4) The rocket is precisely modeled in RockSim 9.0 to confirm stability and the ability of these motors to lift this rocket.

5) CG was confirmed to be proportionally further forward than on the FlisKits Thunderbird.

6) Estes engine chart rates a lifting capacity of 10 oz per D12-5 motor, so if all motors had fired the ejection charges would have fired near apogee.

7) Estes engine chart rates a lifting capacity of 14 oz for the D12-3 motor, so concerns about not flying straight off a 6 ft rail would only come into play in the event of a very high ignition failure rate.

8) An H cluster requires a 200 ft safety distance, which was observed as it always is at CMASS. This creates a very large safe area around the launch for a flight that was expected to reach 480 ft and achieved a perfectly straight flight to about 120 ft.

9) I chose to err on the side of assured stability by using only less heavy D motors for the first flight. This increased the risk of low overall altitude but gained stability margin. Future flights will include some BP E's and an AP F or G as noted previously. I wanted and got a straight up flight off the rod.

Additionally, in over 50 cluster flight firing more than 300 motors, I have averaged well over 95% of loaded motors fired. In more than 30 of those flights I fired all engines. I have never had any rocket ever approach spectators under power, clustered or single-motor.

If every rocket that failed to recover well was judged to be a mistake in the first place, we would all be building model trains...

 

[WillMarchant]

I had been very careful placing and retaining the igniters, and had expected to fire more motors. 3 of the 4 motors not fired showed signs that the igniter had fired but failed to start the motor, the in the fourth it did not appear the igniter fired.

Next time I will:
1) replace some of the Estes D's with E's
2) use more pyrogen when dipping these igniters
2) load a soon to be available AMW/CTI 24mm F or G reload in the center motor tube

Sorry to hear about the less than nominal flight, Boris. I'm curious about the three igniters that appear to have fired but failed to start their motors. Do you think that they pulled out of the motor? Item #2 above makes me think you think that they were in the motor. Or did the igniters malfunction somehow? Thanks for doaing a detailed post-action report, I found it very useful and interesting!

 

[patelldp]

Dave

Boris is an experienced L2, and has flown many flight with up to 30 motors with an extremely high percentages of engines igniting. His Rocksim of the rocket had the apogee at 480' and a peak acceleration of 10 G (the T/W is 11 or 1 per engine) so even with 4 engines out he would have expected to exceed 160' and trigger the altimeter, and there was no wind when he launched.

Most importantly, there was the required 200' separation distance between the rocket and the closest spectators and the launch rod was angled away from the spectators.

Rockets do occasionally fail to operate properly and that's why we have range safety procedures to handle the failures. The required separation distances and angling the rod away from the spectators toward a cleared ballistic recovery zone are two proven methods to mitigate flight failures and insures that no one will be injured in the event of a failure.

I was not the RSO but IMO, the planning of this flight and the launch pad location met all reasonable criteria for oking the launch. The ballistic impact zone was clear so if there was launch or recovery failure the rocket would land in the designated ballistic impact zone. And that's what happened.

Bob Krech, CMASS Senior Advisor

Rob,

I am also an experienced L2 rocketeer. My original post was not really meant to focus on the safety of the rocket, the idea of the rocket reaching the flight line was more or less an afterthought. I have heard good things regarding the safety of CMASS as a club and have no quarrels there.

That being said, I have never heard of anything of that size flying that low before at a club, especially a cluster. I could draw up a rocket right now that could fly to 400' on a 54-2550, but I cannot imagine the fine folks at METRA (where I fly) allowing me to do that, it's just not that good of an idea. I remember a thread a while back that mentions people flying excessively large rockets on small K motors to low altitudes and their altimeters failed to arm, as well. The people involved got a good stern talking to on the forums, why not this scenario as well?

In terms of the cluster, he has a good track record, but that doesn't necessarily mean it's going to work, as is the case here. This flight should allow him to reassess his methods, potentially implementing a flash pan ignition system that, from my experience, is slightly more successful.

I could be wrong, but it just seems borderline at best to me. Could be the results, I guess I have the easy job of just bringing attention to what has already happened, whereas you and he have to justify why it got off the ground.

 

[bobkrech]

Rob,

....That being said, I have never heard of anything of that size flying that low before at a club, especially a cluster. I could draw up a rocket right now that could fly to 400' on a 54-2550, but I cannot imagine the fine folks at METRA (where I fly) allowing me to do that, it's just not that good of an idea. I remember a thread a while back that mentions people flying excessively large rockets on small K motors to low altitudes and their altimeters failed to arm, as well. The people involved got a good stern talking to on the forums, why not this scenario as well?...

Why is it more unsafe to launch a large rocket to an apogee of 400' than to 1000' or higher?

When you run 6-DOF Monte Carlo simulations to determine where a rocket will end up with a parachute recovery and a ballistic recovery, it's clear the ballistic impact zones and the safe recovery zones are far smaller for a rocket launched to a planned apogee of 400' than to 1000'. In fact to first order, it's the ratio of the altitudes squared or 6.25 times smaller for a rocket with a 400' apogee than a rocket with a 1000' apogee.

The minimum separation distance for a K-motor is 200' . If a rocket is designed to apogee at 400', and the launch rod is angled 5 degrees off vertical, the center of the ballistic zone is 150' downrange. If you allow for a +/- 5 degree tip-off error, this predicts a 150' radius circle as the probable ballistic impact zone centered 150' downrange, with possible ballistic impacts as close as the pad or as far as 300' from the pad, and 200'-500' from the spectators.

If one were to launch a high power rocket straight up (which is never done at professional test ranges with unguided rockets because the directionality is not controlled), it's trajectory can go in any direction. But still, if launched to a design apogee of 400', it is still unlike to have a ballistic impact in the spectator area, since the probable impact area is within 150' of the pad and at least 50' away from the spectator area.

On the other hand, a rocket with a designed apogee of 1000' has a probable ballistic impact zone radius of 375'. If the launcher is not angled 5 degrees off vertical away from the spectator area, it's impact zone will rapidly move toward the spectator area. If launched vertical, the probable ballistic impact zone extends into the spectator area by 175' if the spectators are at the minimum 200' separation distance, and the overlap increases with higher apogees.

The safe recovery zone also increase with the square of the altitude. So while is may seem that a higher planned apogee is safer, the opposite is true because of the larger impact and recovery zones associate with higher flights.

My personal preference for a low altitude high power flights is a break-wire launch detection scheme with a barometric apogee detector, and ideally a timer set to 2 seconds after the planned apogee. In most altimeters this arms the barometric recovery system 0.5 seconds after lift-off regardless of altitude and should be able to detect a 400' apogee within 2 seconds or less. A properly designed L2 main chute should be extended in 0.5 seconds nominally and 1 second at the maximum. If the charges fire at apogee, the chute should be fully extended after the rocket drops -4' and is traveling at -16 ft/s. If it's 0.5 seconds late, the drop is -16' and the velocity is -32 ft/s, and the chute should be fully inflated.

In a dual deployment scenario, the rocket is typically free falling at 60 ft/s, so during a nominal 0.5 second extension time, the rocket drops only 30', and the chute should inflate very quickly. That's roughly equivalent to a 2 second late deployment. A rocket that deploy 2 seconds after apogee has dropped only -100' before the chute inflates, and if 3 seconds late, it has dropped -200'. If the planned apogee was 400', then for the latter case, the rocket would be on the ground under chute 10 to 15 seconds later. If the chute stripped, the rocket impact the ground 2 seconds later. Regardless, the uncertainty in the LZ is small, and if angled away from the spectators, it should not drift back to them.

The same apogee drops apply if the apogee is 1000' but the time in the air is proportionally greater, so the LZ is proportionally larger. Because the time in the air is greater for an apogee deployment, and if there is a breeze, the LZ could easily extend into the spectator area if the launch rods were not angled away from the spectator area and in a direction that wind drift did not bring the LZ into the spectator area.

As CMASS Senior Adviser for the last 6 years, I'm responsible for safety at CMASS launches. As a experienced professional working with hazardous and/or explosive materials for almost 4 decades, I'll gladly compare our practices with other professional facilities, not to mention other hobby groups. Given the two options, I'll choose low and slow, and I know where they'll go, over fast and high, and who knows where they'll land, which is what I saw at METRA the one time I launched there.

At CMASS launches, our pads are angled away from the spectators, and depending on wind direction, the splash zone is either a swamp to the SW, or the 3000' open field to the east. In 2005 we adopted a launch policy similar to Ted Cochran's Launch Safe (about a month before it came out), and developed a highly experienced and skilled launch crew that takes safety seriously. Since then, we have not had a ballistic impact outside the field, and only a handful of high power rockets outside the intended splash zones. Only 2 of ~500-600 high power rockets have landed outside the field, and both were due to main chute deployments at apogee.

Conducting safe launches requires constant vigilance and continual reminders as our small, self-regulated hobby is only one major accident away from outside regulation. We determine the probable cause of all high power flight failures, and discuss our findings with the flier to prevent it from happening again, and at least at CMASS, the system appears to be working.

Bob

 

[patelldp]

Well, seeing as I never ran any Monte Carlo simulations with 1 D.O.F., much less 6, I will stand corrected on this one. Please delete my posts, as I was surely flawed in my logic.

I hope you get this rocket to work.

 

[delta22]

Will,

I'm not sure of the root cause for the ignition failures in this flight.

In 2007, using Rocketflite ML igniters, I fired 76 of 77 motors attempted in clusters. In 2008, using Rocketflite MF ematches, I had a 99% success rate. In 2009, using the same MF ematches my success rate dropped to just over 90%.

Humidity was very high at many launches last year, more than once I had issues getting tape to stay in place while holding igniters in the motor nozzles.

Last week I static test fired a series of Estes D motors using MF ematches. Every time I positioned the ematch 1/4" away from the propellant the motors failed to fire. Retaining the ematches against the propellant with either the small wad of tissue, or the tight masking tape pictured, or both, resulted in firing motors 12 times in a row without fail.

I thought I had identified and resolved the primary cause of the ignition failures seen last year, and was very disappointed to have so many ignition failures in this Thunderbird flight. I am confident that the all ematches were tightly contained against the BP propellant.

The next flight I did was 7x D12-0 cluster from the same pad and all motors fired. For this flight I used exactly the same method of igniter retention, but used the longer burning Rocketflite ML igniters.

The ML's were all wired in parallel. The MF's were wired in series strings 3 ematches long, which were then connected together in parallel. This allows the MF's to fire using about 7x less current than the ML's require. Detailed description of this method at:
https://www.bpasa.com/Rocketflite.htm

Past tests have demonstrated that the MF ematches start BP motor ignition in about 0.1 second. This was measured by counting video frames in a series of static and flight tests that positioned an LED in the field of camera view that was connected to the ignition circuit.

Any suggestions would be appreciated.

 

[dixontj93060]

Boris,

Plano-Doug which doesn't visit here too often is a master at clustered BP flights. You might PM him on RP for some input.

-Tim

Will,

I'm not sure of the root cause for the ignition failures in this flight.

In 2007, using Rocketflite ML igniters, I fired 76 of 77 motors attempted in clusters. In 2008, using Rocketflite MF ematches, I had a 99% success rate. In 2009, using the same MF ematches my success rate dropped to just over 90%.

Humidity was very high at many launches last year, more than once I had issues getting tape to stay in place while holding igniters in the motor nozzles.

Last week I static test fired a series of Estes D motors using MF ematches. Every time I positioned the ematch 1/4" away from the propellant the motors failed to fire. Retaining the ematches against the propellant with either the small wad of tissue, or the tight masking tape pictured, or both, resulted in firing motors 12 times in a row without fail.

I thought I had identified and resolved the primary cause of the ignition failures seen last year, and was very disappointed to have so many ignition failures in this Thunderbird flight. I am confident that the all ematches were tightly contained against the BP propellant.

The next flight I did was 7x D12-0 cluster from the same pad and all motors fired. For this flight I used exactly the same method of igniter retention, but used the longer burning Rocketflite ML igniters.

The ML's were all wired in parallel. The MF's were wired in series strings 3 ematches long, which were then connected together in parallel. This allows the MF's to fire using about 7x less current than the ML's require. Detailed description of this method at:
https://www.bpasa.com/Rocketflite.htm

Past tests have demonstrated that the MF ematches start BP motor ignition in about 0.1 second. This was measured by counting video frames in a series of static and flight tests that positioned an LED in the field of camera view that was connected to the ignition circuit.

Any suggestions would be appreciated.

 

[WillMarchant]

Were the non-firing motors in that central cluster? It looks like the tape covered all those motors in a single mass. I've often seen cluster igniters fire in a pattern depending upon the arrangement of the wiring and the current capacity of the firing system. I'm wondering (strictly a guess) if a couple of those motors came up to pressure and started moving the tape mass which might have dislodged the other igniters?

If you still have the wires from the igniters you could make up another set of igniters and see if they fire in a pattern...

 

[delta22]

Will,

Three of the unfired motors were in the central cluster. One was in an outboard booster. Masking tape was still loosely positioned over two of the unfired motors. Unfired motors were distributed one on each series string of ematches.

Both wadded tissue and tape retained each ematch, and each ematch exited all retention completely at the edge of it's motor.

There well may be something different happening at the field electrically than occurs at my test setup at home. Both use 23 Ah gel cell batteries and relays. After the flight I checked the battery at the field and measured 12.9V under no load. That same battery and relay successfully fired 7x D motors using ML igniters in parallel, a much higher current draw.

The leads at the field looked like 16 awg and about 20 ft from battery through relay to pad.

The next couple CMASS launches are at smaller fields. I plan to experiment further with small clusters to try to determine what is happening differently on the field from static tests at home.

Expanding the list of things I am considering doing differently next time:
> more slack in pad igniter leads - more time for igniters
> reduce length series strings of MF's to 2 - increasing voltage drop across each MF possibly making firing time more uniform and fast in the face of voltage droop
> add more pyrogen - more heat
> more total engine impulse

 

[delta22]

Bill S.,

The suggestion of an accelerometer based deployment is interesting, it certainly would have helped with this flight's deployment.

In the event of a gradually arcing flight, a barometric altimeter would do better.

I guess a flight computer that did both would be optimum.

Bob K.,

Thanks for the fascinating and detailed analysis.

Tim,

Thanks for the contact.

 

[delta22]

The Thunderbird has been repaired.

One of the boosters was re-attached. One was epoxied to repair the tear in its center.

All were filleted to the booster to tighten the bonds. At the same time fine tuned fin alignment.

Then re-painted the body.

Will fly at the CMASS Amesbury launch 7/17/10 on 11 motor cluster:
7x Estes D12
3x Estes E9
1x ProX F30 (24mm 3gr) White

Using high output cluster box described in this thread:
https://www.rocketryforum.com/showthread.php?t=11766

to fire Rocketflite ML igniters wired in series strings of 2 or 3 which are then attached in parallel. Details on this method in the website in my signature.

The cluster box will push about 40A at 15V into this igniter load, which should get the job done.

 

[jflis]

Saturday is looking better and BETTER!

 

[delta22]

At the 7/17/10 CMASS launch, the Thunderbird flew perfectly to 643ft.

Fired all 11 engines and had a clean deployment and recovery.

The central Prox F30 had a nice long burn and pleasing white fireball. Surrounded by 3x E9 and 7x D12 the fire and smoke effect worked nicely for a liftoff that was fast enough to be safe off the rod, but long and slow enough to enjoy.

 

[delta22]

Used Rocketflite ML igniters to light the BP motors and the supplied CTI ematch with the ProX motor.

The MLs were wired in series strings of 2 or 3 to reduce current load.

Used the cluster box just completed for power, discussed here: https://www.rocketryforum.com/showthread.php?t=11766

 

[jflis]

Man, this thing was SWEET!

Here are some pix:

• On the pad, getting ready for flight
• LIFTOFF!
• Boost
• Recovery

It was a very, very, very good flight

 

[RangerStl]

The Thunderbird flew perfectly Saturday to 643ft.

Fired all 11 engines and had a clean deployment and recovery.

The central Prox F30 had a nice long burn and pleasing white fireball. Surrounded by 3x E9 and 7x D12 the fire and smoke effect worked nicely for a liftoff that was fast enough to be safe off the rod, but long and slow enough to enjoy.

Have you determined what the ignition issue at the last launch was caused by?

 

[rockie]

Grats on succesfull launch ! Very cool

 

[delta22]

Man, this thing was SWEET!

It was a very, very, very good flight

Thanks Jim.

Another great rocket inspired by a cool FlisKits kit. When I saw the FlisKits Thunderbird, I knew one was in my future.

Have you determined what the ignition issue at the last launch was caused by?

I believe it was a combination of factors, making it a challenge to resolve.
1) Voltage sag in the club launch system. Good battery and relay, but 15-20ft of 16awg leads cost several volts under heavy load.
2) Contributing to staggered firing of ematches and/or some failures to fire.
3) Humid environment making it harder for the ematches to quickly and reliably fire motors.

Resolution was high output electrical source (= cluster box) capable of firing many very hot, longer burning Rocketflite ML igniters all at once. Estimate that the cluster box pushed 600-700W (40A+ at 15V+) into the igniters to kick this cleanly off the ground.

 

[troj]

Why is it more unsafe to launch a large rocket to an apogee of 400' than to 1000' or higher?

When you run 6-DOF Monte Carlo simulations to determine where a rocket will end up with a parachute recovery and a ballistic recovery, it's clear the ballistic impact zones and the safe recovery zones are far smaller for a rocket launched to a planned apogee of 400' than to 1000'. In fact to first order, it's the ratio of the altitudes squared or 6.25 times smaller for a rocket with a 400' apogee than a rocket with a 1000' apogee.

Something else to consider, that hasn't been discussed in any part of this, is the amount of time required for a parachute to inflate. The bigger the parachute, the longer it takes. Also, how the parachute is packed has a bearing on this, as well.

-Kevin

 

[delta22]

At the CMASS 9/18/10 launch the Thunderbird went up again for another great flight.

The central smokey motor put some black smoke in the middle of the cluster and the long burning Es were a nice effect.

Fired all 11 motors and went up 776 feet. Using Rocketflite ML igniters and cluster box: https://www.rocketryforum.com/showthread.php?t=11766

Loaded with:
1x ProX F79 Smokey Sam
4x D11-P
6x E9-6

 

[spacecadet]

Outstanding and a fitting tribute to British engineering, thankyou.
I'm surprised Adrian Hurt hasn't chipped in here.
His Thunderbird is much smaller (D and Cs) and the launch is truly alarming to small children but perfectly safe, naturally. The boosters separate at fairly low altitude so it's all plainly visible.
Quick match does the job there.

 

[delta22]

Outstanding and a fitting tribute to British engineering, thankyou.
I'm surprised Adrian Hurt hasn't chipped in here.
His Thunderbird is much smaller (D and Cs) and the launch is truly alarming to small children but perfectly safe, naturally. The boosters separate at fairly low altitude so it's all plainly visible.
Quick match does the job there.

Thank you and thanks to the RAF for coming up with one of the coolest looking rockets of all time.

When I first saw the FlisKits Thunderbird kit I thought is was a very cool looking rocket, when Jim told me it was a model of a real British anti-aircraft missile I had to build one.

I chose to go for maximum launch fire and smoke effect, and not deal with the additional complexity of separating boosters.

 

[delta22]

The CMASS 11/21/10 launch was the last of the season. The British Electric Thunderbird went up for the fourth time in cool (35 F), sunny, low wind conditions.

After the previous flight I had observed some burned cardboard inside the main body tube. The combination of 7x motor ejection charges, even with individual baffles, plus a 1.0g altimeter fired ejection charge was taking its toll.

To protect the longevity of the rocket I painted the inside of the body tube and tops of the baffles with two coats of thinned epoxy. Used West 206 with 25% alcohol added by weight.

During original construction the 7x engine tubes and all baffle surfaces were coated similarly, but had not done the inside of the main body tube until now.

Pictures:
My son's friend contemplating anti-aircraft missiles as a source of entertainment.
Rocketflite MF igniters wired in a parallel and series combination to balance voltage vs. current ignition requirements.

 

[delta22]

The flight went perfectly, firing all 11 motors and reaching 860 feet.

With low winds on the field, loaded a G "White Longburn" in the center motor position, creating a nice fireball effect.

Motors loaded:
1x ProX 24mm 6grain G65 White Longburn
6x D12-5
4x D11-P

I definitely like the ProX White Longburn motors, the bright, well formed fireball is a great effect. The F30 and G65 are some sweet motors.

 

[delta22]

One of the booster tubes has started to soften/delaminate very slightly where the bottom of one of the big fins attach to it. Likely from impacting the ground on recovery. (Not visible in picture)

Will drill tiny holes into the area from the bottom and fill with thin, slow, West 206 epoxy.

Other than this, the rocket is in great shape and will fly again when launches pick up again in the spring.

 

[delta22]

Flown 3 times in 2011, firing all 33 motors loaded:

[video=youtube_share;IGOnMtm7wL4]https://youtu.be/IGOnMtm7wL4[/video]


flight 5: 4/30/11 CMASS Amesbury, MA
1x F30 + 6x D12-5 + 4x D11-P:

 

[delta22]

flight 6: 11/5/11 CMASS Amesbury, MA

1x F30 + 6x E9-6 + 4x D12

The 1x F30 + 6x E9 engines in the center give a nice long 2.8 second burn and typically hit about 700 ft altitude.
Nice long fire and smoke flight effect and very recoverable for this single deployment rocket.

 

[delta22]

flight 7: 11/19/11 CMASS Amesbury, MA

1x F30 + 6x E9-6 + 4x D12

Thanks to Dave Lindbergh for these pictures.

First picture shows me connecting igniters to the cluster box I have been using for all clustered flights since early 2010.
Thread: https://www.rocketryforum.com/showthread.php?11766-Cluster-Launch-Power

 

[delta22]

Flown twice in 2012, firing all 22 motors loaded.

Both flights on 1x F30 + 6x E9-6 + 4x D11-P, a great motor combination for this rocket, a nice long burn pushing the rocket up 500 to 800 feet.

The English Electric Thunderbird is a very cool looking surface to air missile deployed by the British from 1959 to 1977, with four drop off boosters with canted nose cones and engine nozzles.

This is a 3/20x scale model of the actual missile and a 2x upscale the FlisKits kit.

As built: 11x motor cluster of 24mm engines, 42.5 inches tall, 7.5 lbs flight weight.

[video=youtube_share;3BcS1bDSPw8]https://youtu.be/3BcS1bDSPw8[/video]

flight 8: 10/6/12 CMASS / MMMS Berwick, ME

 

[delta22]

flight 9: 10/20/12 CMASS Amesbury, MA