SpaceX Falcon Heavy Build 4.0 (1:65 Scale, BT-70 tubes) - a work in progress...

[boatgeek]

Yeah, I've read quite a bit on the flash pan approach. One comment stood out: that the flash pan technique is "hard on rockets". This Falcon Heavy would require three flash pans or a really big flash pan. I'm really trying to find a new way to do clusters. As I mentioned, the 9 cluster Falcon 9 launches using quick fuse have been very reliable. So this Falcon Heavy is really just a 3 igniter problem. I have had problems at club launches where the batteries were old and my clusters worked when I used my personal "jump start" lithium ion battery. I'm going to explore that variable this weekend with some test fires.

I would strongly consider using some of the MJG ematches at least for the center motors of each core. They light reliably and with relatively little current, so you should get ignition on the main motors at the same time. The long leads would also help connect stuff at the pad.

 

[jmasterj]

I would strongly consider using some of the MJG ematches at least for the center motors of each core. They light reliably and with relatively little current, so you should get ignition on the main motors at the same time. The long leads would also help connect stuff at the pad.

I've done up to 4 of these with no problems, and seen pictures of up to 7. I don't know if they fit in A10 nozzles. Worth investigating for sure!

 

[bad_idea]

I've done up to 4 of these with no problems, and seen pictures of up to 7. I don't know if they fit in A10 nozzles. Worth investigating for sure!

They fit A10s, and A3s too.

Edit to add: a spider would drop the cost per flight over a whole mess of MJG starters at 80 cents apiece, but the MJGs sure save on the engineering, which is why I've pursued spiders myself.

 

[MaxxQ64]

Making an assumption here, but since flash pans are "legal" for cluster ignitions, might this method also work for large clusters (20 motors for this project):

https://www.rocketryforum.com/threads/soyuz-1-25-why-not.156583/page-5#post-2496916
Starts a few posts down on page 5.

Might need to find some much thinner spaghetti for A10's...

 

[Switch]

The quick fuse method I've been experimenting with is working great on my Falcon 9 with ~99% ignition in the last 15 flights. It's basically a version of the spider method & the clubs I've launched with have not had a problem with it.

So I'm just dealing with getting three igniters to light on the Heavy, which shouldn't be that hard. For my launch tomorrow, I'm going to test several motors with the clip whip before trying the Falcon Heavy. The mistake I made last week was just going for it without any testing before hand. And I agree that I should not be using Estes igniters on the Heavy.

 

[Switch]

Finally 27 motors ignited!


There is this unfortunate detail about the E12 in the core exploding.


Here you can see the core motor partially ejected with the upper booster mount shearing on the left booster.


The explosion of the core E12 caused the 2nd stage to pop out like a Champaign cork,


And the A10PT motors in the core carried it off to places unknown.

I had an E12 fail with similar results in my Falcon 9 last year.

 

[boatgeek]

Finally 27 motors ignited!


There is this unfortunate detail about the E12 in the core exploding.


Here you can see the core motor partially ejected with the upper booster mount shearing on the left booster.
View attachment 614907

The explosion of the core E12 caused the 2nd stage to pop out like a Champaign cork,
View attachment 614913

And the A10PT motors in the core carried it off to places unknown.
I had an E12 fail with similar results in my Falcon 9 last year.

Nooooooooo!

 

[Switch]

Well crap. I don't think the core E12 ignited after all. The only thing holding the core motor in place is the magnetic shear pins for the landing legs. On further analysis, I think the booster D12 motors ignited but not the core E12, causing the core to eject due to Newton's 1st Law. The quick fuse in the core ignited from the flame of the boosters and then the E12 ignited. But by then the core recovery chute had snagged on the launch pad, causing the core and remaining booster to recoil back & release the 2nd stage. Because I lost the core 9 cluster motor mount, I have not way to be sure...

Here you can see the reason the 2nd stage ejected



On the upside reverse recovery of the one booster that tumbled into the air worked and the damaged to the Falcon Heavy parts was minimal.

 

[bad_idea]

Cluster flights lend themselves to complex forensics when things go awry. Photos and videos help a lot, and it seems you have some good photos to go by when seeking root cause.

 

[Switch]

Date: 12/16/2023 13:35
Location: Snow Ranch
Event: LUNAR Club Launch
Photo credits: Rob Heil

 

[Switch]

I was hoping to get a video of that last flight so I was waiting to post this commentary. But that didn't happen so I'm just going to post these notes:

I sure wish that booster separation was intentional. It almost looks like those boosters are turning around to fly back to Kennedy Space Center. But no, they fell off as the rocket tumbled shortly after takeoff.

On the positive side, I got this bird off the pad for the first time, all of the 1st stage recovery worked, and 2nd stage ignition worked! Also on the positive side, 24 motors in the 1st stage ignited!

But failure analysis is hard and humbling. And that is the task at hand...

The rocket clearly tumbled on the Y axis*. An obvious modification would be two lateral fins on the boosters. Participants at the launch also suggested a longer launch rod.

I think the reason some motors are not igniting is that the igniter and igniter plug in the center D12 and E12 motors is impacting the fuse mess that is looped under the center motor.

* I'm assuming Z is up, X is left/right on the line of the boosters, which makes Y on the front/back axis of the rocket the launch rail)

 

[Switch]

It finally worked!

(mostly)...

So with the results of test flight #4, I set about adding fins to the boosters and lengthening the launch rail. Another lesson learned from my previous three flights is that I needed to bring back the Transport, Elevate, & Launch (T.E.L) system, also known as the "strong back"



I repurposed the T.E.L that I used on test flight #1 & added mounts for the 2nd stage. This allowed me to transport the rocket without having to remove the landing legs that had fins attached. This also provides a base for the 1st stage to rest on when placed on the launch pad, so it is not just hanging from the launch lug.


The T.E.L is a 4 foot long 1" launch rail. The plan was to attached my 6' rail so I would have a 10 foot rail to launch on.

You can also see in the above photo, the addition of 4 more fins on the boosters and my new method for running the quick fuse to it does not cross directly under the igniter plug. I used wire on the core before deciding it would be easier to use super glue, which I did on the boosters (the super glue turns the fuse red).


Date: 01/27/2024
Time: 12:42
Location: Snow Ranch, California
In flight photo credits: Rob Heil






In the above photo, I still needed to adjust the boosters to mate withe lower booster mounts of the core rocket. But note that the LOX fuel lines double as the recovery cord. The cord loops around a M2 screw and is attached to the motor by yellow Kevlar™ string.






Lift off was almost perfect. But only the center motor on the right booster ignited plus 4 motors in the left booster ignited and 6 in the core. Fortunately the D12-0s in the boosters and E12-0 in the core all ignited. So that change to my layout of the fuses to the A10PT motors proved to be a bad move.



I think the turn here is due to the right booster having only one motor burning. Most importantly, the tumble that occurred in December was fixed by adding fins to the boosters.




Separation of the boosters at BECO worked, but the motor on the left booster did not fully eject. You can see it is bent at the bottom of the booster. The reverse recovery on the right booster worked perfectly (in this photo the chute had not yet deployed). The red parts on the lower left are the shrouds that protect the parachutes and were designed to fall freely. If you look carefully you will notice some of the tubes inside the landing legs separated from the body. That's a known bug that has been fixed (I think) in newer versions of the landing leg tubes. Also, you can see that one of the landing legs on the core has deployed prematurely.






The ignition of the 2nd stage worked and reverse recovery of the core and 2nd stage recovery also worked. The red Tesla Roadster was also recovered.




This is the most successful launch yet so I am quite pleased.



PS: The core rocket landed in the creek and I thought is would be a total loss since it is just adhesive paper wrapped around a BT-70 tube. But it dried out and will fly again!

 

[ReynoldsSlumber]

That's beyond spectacular! So many distinct elements...

 

[Switch]

Date: 03/09/2023
Location: Snow Ranch
Event: LUNAR Club Launch
Photo credits: Rob Heil

I flew the Falcon Heavy two times that day. On the the first launch, the second stage motor did not ignite but 26 out of 27 1st stage motors ignited and recovery of all 1st stage parts was successful.

The second launch also had 26 out of 27 1st stage motors ignite, but recovery of one booster failed while everything else worked mostly as planned.







Problem with landing legs deploying prematurely..



On the 1st launch, the boosters separated from the core prior to BECO, but it looked awesome!


Core continues to fly after BECO



2nd stage continues after MECO


2nd stage recovery in process with red Tesla Roadster deployment

 

[Donnager]

Here's my update on this build. I've got everything ready for assembly (I think):
View attachment 554688

Firs step is to test fit everything. Starting with the magnets for the lower booster attachments:
View attachment 554689
View attachment 554690
View attachment 554691
When I did the test fit for the BT-70 couplers, I realized I forgot to included a sleeve at the mid point to hold the coupler in place. Recall that one goal of this project was to minimize the use of glue & use the adhesive paper skins to hold the body tube parts together. So I need to reprint these couplers with a BT-70 tube OD ring to old them in place.
View attachment 554692
meanwhile the engine assembly starts with taking the two top magnet centering rings and gluing them together:
View attachment 554693
The magnet mount bore runs all the way through to aid adjustment and ejection of magnets & tune the number of magnets required.
View attachment 554694
View attachment 554695
Here, the two parts are glued together with superglue to form one part holding magnets with opposite polarity in place for upper booster attachment and separation:
View attachment 554696

Holy hell.

You should see my work area (and the utter lack of layouts, other than in a bag near my worktable).

This is making me self conscious.

 

[Switch]

Also just got this photo of the Falcon Heavy launch from another club member. His rocket was coming down as mine was going up...

 

[ReynoldsSlumber]

Also just got this photo of the Falcon Heavy launch from another club member. His rocket was coming down as mine was going up...

OK now that is super cool. (In the days of yore I made a 35mm film camera rocket, and my dream was to get an aerial photo of another rocket taking off.)

 

[Switch]

After a great launch in January and a tragic CATO in February, I'm learning the value of getting quality video to document my successes and failures. It is really frustrating to have problems that I don't have a record of. It makes is really difficult to reconstruct what went right and what went wrong. And I'm also learning how to edit video to better represent my progress and setbacks. Just getting the videos was a lot of work on this last launch. I staged multiple cameras for this launch. I found that old iPhones can be used to capture video and here is my first attempt to stich multiple videos with still photos to try to tell a coherent story about a particular launch.

 

[Switch]

So after a fun launch there's the not-so-fun damage assessment and repairs. And the fun challenge of figuring out how to improve parts for the next launch.

The 9-cluster motor mounts have an obvious weak point: they are snaping at the end of the BT-20 motor tube where the 3D-printed part has holes to reduce mass. The magnet rings that hold the landing legs in place are also taking a beating.


The 3D printed parts are easily replaced and I can remove the holes to make them stronger.



The the loss of the 18" BT-20 tube is really annoying.



Recall that the entire booster motor assembly looks like this:


It runs the entire length of the booster so that the magnetic upper booster attachment to the core rocket can release when the entire motor assembly ejects out the back of the booster during rear ejection recovery.

An obvious improvement would be to swap the upper and lower sections of the BT-20 motor tube so that the shorter section is in the location prone to damage. But I intend to delete the entire 18" BT-20 motor tube by redesigning how the upper booster attachment mechanism works.

Currently a pair of magnets at the upper booster and lower booster attachment secure the booster to the core using the motor assembly.


Shown here with the booster body removed.

The booster ejection charge causes the entire motor assembly with its magnetic sheer pins to slide downward; replacing the attracting magnets in the upper and lower booster mounts with a repulsive magnets that push the booster away from the core.


Shown here with the booster nose cone in place for reference.

The booster motor then needs to completely eject from the booster to deploy the recovery chutes. Any sideways force on the BT-20 tube as it ejects from he booster body damages the BT-20 tube.


Here an otherwise successful launch has a booster assembly with a failed BT-20 tube.

My solution is to create a magnetic booster attachment that uses the pressure of the booster motor ejection charge to push the booster attachment magnetic sheer pin upward instead of using a BT-20 tube to pull it downward.

This means creating a part that moves up and down inside the booster nose cone, like a piston in an automobile engine.


Here the new booster magnetic sheer pin is shown in orange inside the booster nose cone.

The new mount would use a centering ring to support two magnets: the first attracting magnetic sheer pin would hold the booster to the core.




A 2nd new part (in blue above) will hold the booster attachment part in place and includes a guide to prevent it from rotating inside the booster nose cone.


The pressure from the ejection charge will move the booster magnetic mount upward; replacing the attracting magnet with a repulsive magnet.


Air pressure inside the booster nose cone will escape through the holes hidden behind the grid fins.


The completed assembly looks like this:

Viewed from the side: attracting magnetic sheer pin in place.


Viewed from the bottom with the booster attachment in the lower position.

The pressure from the ejection charge moves the magnetic sheer pin booster mount upward inside the booster nose cone.

The guide that prevents the booster attachment from rotating or falling down inside the booster is friction fitted to allow tuning of the position of the booster attachment magnets.

So this is how I hope to delete the VT-20 motor tube.

The new attachment magnetic mount inside the booster nose cone.

In summary, the new magnetic booster attachment mount:

1. Weighs less than the old BT-20 based mechanism
2. Reduces the total number of upper booster parts by 1: 4 instead of 5 or 20%
3. Eliminates a single point of failure: the booster motor assembly 18"BT-20 tube
4. Eliminates the need to align the upper and lower booster magnetic sheer pins to the core.
5. Is backward compatible with existing booster nose cones.
6. Simplifies installation of the booster motor, because it is so much shorter
7. Increased probability of booster rear ejection success, because the motor assembly is so much shorter
8. Moves the center of gravity upward by concentrating the parts in the nose cone, rather than along the length of the booster.

I just don't know if it will work...

 

[Switch]

I was reviewing video from flight test #6 and noticed that I got a record of the booster ejection charge gas escaping through the holes in the nose booster cones. I had not realized how violent the ejection charge was or how much gas was escaping & possibly not contributing to the rear ejection of the motor assembly. Here, the boosters separated from the core prematurely and tumbled. So my new booster attachment mechanism should result in a more powerful motor ejection and the shorter motor length help even more.

View attachment Falcon Heavy 02242024 Closeup of booster ejection.mp4

 

[Switch]

Since my last post. I've flown the Falcon Heavy 8 more times! That's four 500 mile round trips from San Francisco to Reno Nevada in six weeks. In between launches, printing broken parts and putting rockets back together has consumed all of my spare time.

While I continue to iterate on the design, the most exciting news is that the ignition of all 27 motors in the first stage has been successful for the last 5 consecutive flights! And the 3 before that all had 25 or 26 motors ignite. So I think I have finally got a handle on 27 motor cluster ignition. I'm also learning a lot about doing videos of these launches and learning that post production is as much work as building these rockets in the first place.

Here is my 2nd build of the Falcon Heavy core with two redesigned boosters using the new pneumatic upper booster release mechanism and vinyl skins (instead of paper). The new booster release mechanism works, so good bye bent booster motor tubes!


Here you can see most of the motors ignited. As my photography skills improve, I'm learning a lot more about the events that occur at take off.





On the video above, you can see how even as the rocket leaves the pad and launch rail, not all of the motors were ignited but eventually they all do.


I've also solved the fin attachment problem by using clear Elmer's glue. The trick is to apply consecutive layers to fillet the fin attachment (the main point in not using superglue is to be able to reuse the fin or landing leg if the other part is damaged).


Second stage ignition continues to be a challenge. The success rate being something between 50%-75%. I'm pretty sure the second stage is being popped off the 1st stage like a champagne cork before the 2nd stage motor ignites. I'm going to do a series of static tests to debug that. I really should have done that already.

 

[sr205347d]

A key innovation has been the use of quick fuse instead of multiple igniters. This resulted in almost 100% success in lighting all 9 motors in the Falcon 9.

What is your favorite source for that? Is shipping a problem? Is it something you can get at the roadside fireworks dealers?

 

[Switch]

I've been using Bonanza and its not cheap and takes time for shipping. The quality has ranged from great to only OK (as in fuse that is stiff and brand-new looking to really old looking, but that has not appeared to make a difference in practice. In California, there are no road side stands.

https://www.bonanza.com/listings/Fa...reminder&utm_medium=Notifier&utm_source=email

 

[Switch]

Falcon Heavy flight #17.

This one is going to take some time to analyze. After a successful ignition of all 27 motors in the 1st stage, the 2nd stage motor didn't ignite. The video of the flight shows that after BECO, the core rocket motor is not burning. I had an extra long fused to the 2nd stage motor placed deliberately down the center of the main engine motor tube.




A note about the SPACEX logo on the core in the above photo. Back in January the core spent about 20 minutes under water in a small creek at the Snow Ranch LUNAR launch site. The logo ink was water based and mostly disappeared. To my amazement the core itself which is made of PLA plastic, cardboard tube, and self adhesive paper survived. So I printed and applied the SPACEX logo over the original paper skin on the Core.










Notice no exhaust from the core motor. The 2nd stage is ejected passively from Booster Engine Cutoff (BECO)



Something does not make sense and I'm tired. So I will need to look at this some more later.

 

[Switch]

Falcon Heavy flight #18

Location: Artesia Dry Lake, NV
Date: 09/21/24
Time: 12:19




Flight test objectives: I wanted to test the Falcon Heavy with smaller fins that have been working on my Falcon 9. The boosters have the new magnetic/pneumatic upper booster attachment/separation that has only flown a few times. I'm testing some changes to the core motor to improve reliability of the 2nd stage ignition. I'm also testing fin attachment to the landing legs which has been an ongoing problem.

Flight Notes: The smaller fins caused the rocket to tumble just like it did on flight #2 before I added fins to the boosters. 27 out of 27 motors ignited in the 1st stage. Separation of the boosters at Booster Engine Cut-off (BECO) worked. Reverse recovery of both boosters worked and the landing legs were deployed. No fins separated in flight. Ignition of the 2nd stage worked at Main Engine Cut-off (MECO). Reverse recovery of the core rocket failed resulting in the loss of the first 3D printed section of the core, which can be repaired.

With the partial destruction of both Falcon Heavy's in the past 7 days and the complete loss of my only Falcon 9, I'm hard pressed to do repairs before the next launch which is less than 3 weeks away. I've also started a new build that will have all of my latest improvements plus replacing adhesive paper "skins" with adhesive vinyl "skins" which I used on my latest Falcon and were a significant improvement.

 

[Switch]

Falcon Heavy flight #19

Location: Misfits Flat, NV
Date: 10/12/24
Time: 10:50



Flight test objectives: I wanted to test a new design in the core to improve ignition of the 2nd stage. Ever since I made a design change to the core motor of my Falcon Heavy and Falcon 9 rockets, 2nd stage ignition has been failing about 33% . The ejection charge from the E12-0 motor simultaneously ignites the 2nd stage A8-3 motor and ejects the E12-0 motor out the rear of the rocket along with the recovery chutes for both the motor assembly and the 1st stage rocket. The new design adds a BT-50 tube in the center of the BT-70 body tube. The idea is to direct the ejection charge up the length of the BT-50 tube to the 2nd stage A8-3 motor. This new BT-50 tube replaces the function of a BT-20 that used to be attached to the core motor but was causing problems with reverse recovery because the entire 18" BT-20 tube needed to be ejected from the rocket body. The new BT-50 tube remains inside the rocket body and does not eject with the motor.

Flight Notes: The 2nd stage ignition worked. 27 out of 27 motors ignited in the 1st stage. Separation of the boosters at Booster Engine Cut-off (BECO) worked. Reverse recovery of both boosters worked and the landing legs were deployed. No fins separated in flight. Ignition of the 2nd stage worked at Main Engine Cut-off (MECO). Reverse recovery of the core rocket also worked. The only thing that didn't work was recovery of the 2nd stage because the shock cord broke.

A special thanks to: Apogee Rockets (https://www.apogeerockets.com/) for their amazing library of how-to articles and videos and great technical support, without which this project simply would not have happened. And to Sierra Rockets for hosting this event: https://sierrarockets.org/

 

[Switch]

Location: Artesia Dry Lake, NV
Date: 10/20/24
Time: 12:55





Flight test objectives: Same as last flight. All I needed was for everything to work like the previous week and not have the 2nd stage recovery shock cord break.

Flight Notes: Fate had other plans. First stage ignition worked with 26 out of 27 motors ignited in the 1st stage. No fins separated in flight. Separation of the boosters at Booster Engine Cut-off (BECO) worked (sort of). Reverse recovery of both boosters worked but the shock cord on one booster broke. The force of the booster engine cutoff (BECO) caused the 2nd stage to mechanically eject from the core rocket. Ignition of the 2nd stage failed because the 2nd stage ejected before main engine cutoff (MECO). Reverse recovery of the core rocket motor also failed. I have had problems with 2nd stage ignition before and never considered this scenario as a possible cause. Now I'm thinking about in-flight cameras & worrying about the "improvement" I made to booster attachment when I introduced the pneumatic upper booter release mechanism earlier this year.

Next launch is in two weeks (weather permitting). The damage wasn't too bad all things considered. The entire 2nd stage was repaired. Although the nose cone ripped 3/4 down through the BT-80 tube, all I needed was some super glue to get the BT-80 tube looking good as new.



Happy Halloween (Frankenstein build)? It's dead! No! IT'S ALIVE!!

 

[Switch]

This 2nd stage has been through so much. And only thanks to the soft salt surface of Artesia Dry Lake did it survive. That lake bed makes recovery chutes almost unnecessary. But here the nose cone is shoved 3/4 down the BT-80 tube.




So I wanted to bring that 2nd stage back to life. Fortunately I'm using adhesive paper skins:


So sand down the worst of the super glue repair:



And go from this:






To this:





Not perfect but better than no effort..