New Motor certified, April 2023

[Alan Whitmore]

The Tripoli Motor Testing Committee met on Saturday April 29, and certified 1 new motor, the Aerotech N4000W-PS. Total impulse is 18,183.8 N.s, average thrust is 3906 N, and the average burn time is about 4.6 seconds.
This is a reload kit made for the new 98/20480 hardware, and it requires the same sort of gluing procedure for the grains as is used for the O5550X. It's easy to do with 3 people, somewhat harder with 2 people, and impossible with only one person. TMT would strongly recommend having 1 or 2 looks at the youtube video provided which describes the assembly process.
Alan Whitmore
Chair, TMT

 

[jd2cylman]

I’m having trouble finding the N4000W assembly video. Had it once and lost it. Anyone got a link for the search impaired?

 

[jsdemar]

The Tripoli Motor Testing Committee met on Saturday April 29, and certified 1 new motor, the Aerotech N4000W-PS. Total impulse is 18,183.8 N.s, average thrust is 3906 N, and the average burn time is about 4.6 seconds.
This is a reload kit made for the new 98/20480 hardware, and it requires the same sort of gluing procedure for the grains as is used for the O5550X. It's easy to do with 3 people, somewhat harder with 2 people, and impossible with only one person. TMT would strongly recommend having 1 or 2 looks at the youtube video provided which describes the assembly process.
Alan Whitmore
Chair, TMT

Alan,
Could you account for the difference between the two thrust curves?

 

[FredA]

Could you account for the difference between the two thrust curves?

Yes, Please! Or someone from Aerotech.
Not knowing the design margin, but a 30% jump in thrust is usually a similar jump in operating pressure and perhaps nearly a CATO.
What kind of consistency is required? Or inconsistency allowed?

 

[Kallahan11]

Ohh man, you had to go an cert this on my birthday. It's destiny, not this years destiny, but destiny non the less.

Also here's the vid/

 

[Rocketjunkie]

I'm having trouble finding the Elmer's Glue-All Max. With the infrequent need of this adhesive, I need to buy each time. Even in unopened package, it hardens after a couple years. Regular Gorilla Glue does foam too much.

 

[Handeman]

I'm having trouble finding the Elmer's Glue-All Max. With the infrequent need of this adhesive, I need to buy each time. Even in unopened package, it hardens after a couple years. Regular Gorilla Glue does foam too much.

I got mine through Amazon. I couldn't find it local either.

 

[Wayco]

Just bought some yesterday on Amazon, $15 for 8 oz.
Last purchased in early 2021, and it is developing a hard layer, but still the old bottle works.

 

[prfesser]

Glue-All Max is a polyurethane glue, based on isocyanates. It might be worthwhile to store the bottle the same as isocyanates used for propellant: airtight, non-permeable container (glass jar would be better than plastic) and close to room temperature.

Possibly the hardened contents *might* be reliquified by warming gently, up to maybe 110-120 F. Depends on the specific isocyanates in the glue. Some isocyanates dimerize slowly, with two molecules linking together. Warming can separate the dimers.

If the solids are from reaction with moisture, that is not reversible.

 

[Rocketjunkie]

I got some from Amazon and it was *already* too thick to use

 

[Rocketjunkie]

Just bought some yesterday on Amazon, $15 for 8 oz.
Last purchased in early 2021, and it is developing a hard layer, but still the old bottle works.

If I buy 8 oz, I might use 1 oz and the rest would be hard by the time I need it again. Might as well buy 2 at that price and avoid shipping charges but it would *still* be unusable by the time I would need it again. When I need to bond grains I sand the crap out of the liner and use epoxy.

 

[ThirstyBarbarian]

This sounds like an awesome motor.

 

[jsdemar]

Yes, Please! Or someone from Aerotech.
Not knowing the design margin, but a 30% jump in thrust is usually a similar jump in operating pressure and perhaps nearly a CATO.
What kind of consistency is required? Or inconsistency allowed?

It appears to be an unexpected progressive burn, possibly due to glue inhibiting some of the grain faces. Hopefully @Alan Whitmore or @ATGM can elucidate. I haven't seen the instructions, but I'm guessing there's a warning about this.

NFPA 1125 requires a total impulse within a 6.7% SD of the mean, and an average thrust within 20% of labelled value. It may meet the testing requirements, but I think there should be better consistency for a motor of this price. If an experienced TMT person can assemble with this much variation, the student teams at the Spaceport Cup will likely have bigger issues.

 

[FredA]

but I think there should be better consistency for a motor of this price. If an experienced TMT person can assemble with this much variation, the student teams at the Spaceport Cup will likely have bigger issues.

Agree!
Interesting notion of shipping eight grains and gluing them into four.
Seems the whole process as exhibited on the video can result in random levels of inhibition on the grain faces.
There are o-ring spacers, but there is also epoxy everywhere.

 

[mikec]

NFPA 1125 requires a total impulse within a 6.7% SD of the mean, and an average thrust within 20% of labelled value. It may meet the testing requirements, but I think there should be better consistency for a motor of this price.

Agreed. FWIW, a quick-and-dirty TCtracer session shows that burn 1 has an average thrust of 3817 N and total impulse of 18166 Ns, and burn 2 is 3895 N and 18234 Ns, so by those metrics they are pretty close.

 

[swfa]

Unintentional inhibited grain faces should result in lower initial thrust, right? Both burns appear to start at the same thrust level.

 

[jsdemar]

Unintentional inhibited grain faces should result in lower initial thrust, right? Both burns appear to start at the same thrust level.

The initial thrust in a long motor is dominated by the core mass flux and erosivity. The progressive curve after the startup indicates inhibited faces.

 

[Alan Whitmore]

I don't know why there are occasional variations in max thrust in a motor test. I would suppose that SMALL areas of poor bonding of propellant to the casting tube would allow intrusion of surface burn into areas where this should not happen, but any major lack of propellant bonding would result in a quick CATO. It's also possible that pre-pouring variation in mixing efficiency could result in local areas within the propellant of higher or lower local concentrations of burn rate catalyst or metal fuel concentration, or some other variable. I just don't know.
We saw this sort of variation recently when TMT tested the AT J1265ST (relying only on my memory here, forgive me if that's not the right name)
I do know that I see it occasionally in the propellant I mix myself and test on my own private instrumentation. Not often, but enough to make me aware.
John D has pointed out that the NFPA 1125 requirements for motor performance relate to variation in total impulse, delay element timing, and labeled vs measured average thrust. There is no specification for max thrust. I guess, although I don't know, that the gray eminences on the NFPA pyrotechnics and rocketry subcommittee had enough collective experience with rocket motors to realize that this sort of variation in max thrust was just "the nature of the beast", and that the total impulse figures, and the variation thereof, were the more important variable.

Alan

 

[Alan Whitmore]

Agree!
Interesting notion of shipping eight grains and gluing them into four.
Seems the whole process as exhibited on the video can result in random levels of inhibition on the grain faces.
There are o-ring spacers, but there is also epoxy everywhere.

That would be a DOT requirement. Also, 7 grains, not 8. The single grain with no inhibited ends goes at the top. The process of putting epoxy on the grain ends is a "belt and suspenders" approach. Six of the 7 grains come pre-inhibited on one end, and the user just adds more epoxy to the already epoxy coated ends to stick them together. When we built the motors before testing, we tried to get a layer of epoxy on the whole end, but I'm sure there were little spots that had only the pre-applied epoxy burn inhibition.

Alan

 

[jsdemar]

That would be a DOT requirement. Also, 7 grains, not 8. The single grain with no inhibited ends goes at the top. The process of putting epoxy on the grain ends is a "belt and suspenders" approach. Six of the 7 grains come pre-inhibited on one end, and the user just adds more epoxy to the already epoxy coated ends to stick them together. When we built the motors before testing, we tried to get a layer of epoxy on the whole end, but I'm sure there were little spots that had only the pre-applied epoxy burn inhibition.

Alan

Are the grains also glued into the liner? In doing so, some glue can get onto the faces of the non-inhibited grains.

The issue is not a manufacturing problem with bonding to the casting tubes. That would create an initial high thrust and cato at ignition.

The issue happens when the motor is assembled and glue gets onto the as-intended, as-designed, uninhibited grain faces (not the faces which are supposed to be glued together). This will cause a progressive burn (that we see in the higher peak curve). The possible cato would come half way into the burn. Been there done that long ago... this is why there are grain spacers, BTW.

This isn't an issue covered in NFPA 1125 per testing requirements. This is a TMT policy requirement of having complete manufacturer instructions to allow reliable assembly. It takes a lot of juggling to get these motors built, and certainly not by one person alone. If TMT had an assembly issue for 1 of 2 motors, the typically user certainly will, IMO.

If you want to see the affect of a partially inhibited grain on the thrust curve, use Burnsim and inhibit one grain end on one grain of a multi-grain simulation. You will see the same progressive affect.

 

[Alan Whitmore]

. It takes a lot of juggling to get these motors built, and certainly not by one person alone.

Agreed, and stated previously

 

[AeroTech]

It appears to be an unexpected progressive burn, possibly due to glue inhibiting some of the grain faces. Hopefully @Alan Whitmore or @ATGM can elucidate. I haven't seen the instructions, but I'm guessing there's a warning about this.

NFPA 1125 requires a total impulse within a 6.7% SD of the mean, and an average thrust within 20% of labelled value. It may meet the testing requirements, but I think there should be better consistency for a motor of this price. If an experienced TMT person can assemble with this much variation, the student teams at the Spaceport Cup will likely have bigger issues.

We are not sure about the difference. We have occasionally seen this before in other motor certifications. One of the curves was very close in shape to ours. It could be a difference in assembly technique, or something else we haven’t considered. The deviation is gradual and progressive which is not normally indicative of a void or unbond of the propellant from an inhibiting surface. Regardless, the higher pressure is well within the safety margin of the motor hardware.

 

[JohnCoker]

FWIW, a quick-and-dirty TCtracer session shows that burn 1 has an average thrust of 3817 N and total impulse of 18166 Ns, and burn 2 is 3895 N and 18234 Ns, so by those metrics they are pretty close.

Yeah, I'm not convinced higher max thrust makes that much difference in the end.

I don't recall seeing any data about variability of reloadable motors in practice. If anyone has their own test stand, an interesting experiment would be to collect multiple of one motor over a year or two from different sources and compare them. We expect that professionaly-made motors from AeroTech (as opposed to say garage-produced motors from Kosdon back in the day) would have lower variation than NFPA requires, but we don't really know.

The TARC teams do this for certain MPR motors, but those tend to be SUs, which presumably vary less because they can be entirely controlled during production.

 

[JohnCoker]

I wonder if we could do an experiment...

Choose one popular HPR reloadable motor (J maybe) and have various people buy them through their local dealer, build them and deliver them at a national launch. Then test them all and see what variation there is in practice for a "normal" motor. Then we'd have a basis for evaluating the variations in the N4000.

 

[jsdemar]

Yeah, I'm not convinced higher max thrust makes that much difference in the end.

I don't recall seeing any data about variability of reloadable motors in practice. If anyone has their own test stand, an interesting experiment would be to collect multiple of one motor over a year or two from different sources and compare them. We expect that professionaly-made motors from AeroTech (as opposed to say garage-produced motors from Kosdon back in the day) would have lower variation than NFPA requires, but we don't really know.

The TARC teams do this for certain MPR motors, but those tend to be SUs, which presumably vary less because they can be entirely controlled during production.

NFPA 1125 requires the manufacturer to test 1% of each production lot of each motor type, and keep the results on file. If the total impulse of the sample varies more than 10% (among other requirements), the manufacturer is expected to test an additional 2% of the lot. If the results vary again, they're expected to correct the manufacturing error or destroy the lot.

I think the 1% / 2% sample requirement was intended for large volume motor/reload production. I doubt a manufacturer would test any samples from a small lot of large motors.

 

[jsdemar]

We are not sure about the difference. We have occasionally seen this before in other motor certifications. One of the curves was very close in shape to ours. It could be a difference in assembly technique, or something else we haven’t considered. The deviation is gradual and progressive which is not normally indicative of a void or unbond of the propellant from an inhibiting surface. Regardless, the higher pressure is well within the safety margin of the motor hardware.

With the initial thrust the same on the two, it won't be a safety issue for the minimum thrust-to-weight ratio. The final burn-out velocity will be higher for the higher peak curve, but not an issue if the rocket can handle it.

To rule out an assembly issue, you could build the motor purposely using too much glue oozing onto the uninhibited faces. Then take a looks at the test stand data.

Also, depending on the type of glue or inhibitor, and the time it was allowed to cure, the initial startup of the motor could melt the stuff and allow it to spread on the grain faces. This could inhibit more area and cause a progressive burn.

Another way there can be an unexpected hump in the curve is due an unexpected resonance. The low sample rate of the data will make the oscillations look like an "envelope" of the real data. The causes of resonance burning are discussed in the literature. The reasons are limited when it's two motors of the same geometry and propellant.
1) Resonance from the test stand is transferred into the casing. This can be a single tap or pulse at the "right" time which starts the oscillations.
2) There's a pressure threshold over which the combustion is unstable, especially in a long motor, and more so for large oxidizer particles.

I can dig up paper references if needed.

 

[StreuB1]

Are the grains also glued into the liner? In doing so, some glue can get onto the faces of the non-inhibited grains.

The issue is not a manufacturing problem with bonding to the casting tubes. That would create an initial high thrust and cato at ignition.

The issue happens when the motor is assembled and glue gets onto the as-intended, as-designed, uninhibited grain faces (not the faces which are supposed to be glued together). This will cause a progressive burn (that we see in the higher peak curve). The possible cato would come half way into the burn. Been there done that long ago... this is why there are grain spacers, BTW.

This isn't an issue covered in NFPA 1125 per testing requirements. This is a TMT policy requirement of having complete manufacturer instructions to allow reliable assembly. It takes a lot of juggling to get these motors built, and certainly not by one person alone. If TMT had an assembly issue for 1 of 2 motors, the typically user certainly will, IMO.

If you want to see the affect of a partially inhibited grain on the thrust curve, use Burnsim and inhibit one grain end on one grain of a multi-grain simulation. You will see the same progressive affect.

It is due to this that I have adopted the Loki method of grain gluing. Where you build a stack of grains with the spacers, tape the joint, then glue the entire stack at once into the liner. This way, you have zero chance of getting inhibition of the inter-grain sections of the stack. It's also a LOT cleaner to perform, you just need to make sure you dry-fit the entire taped stack BEFORE you glue to make sure the little added thickness at the tape joints does not get in the way during the stack to liner assembly. I have used this method every time and it has not had an issue.

 

[Syclone]

I find it interesting there was a similar thrust curve with the O5500 test 2, but to a lesser extent.

 

[Neutronium95]

I find it interesting there was a similar thrust curve with the O5500 test 2, but to a lesser extent.

The J1265 is another motor that had an anomalous certification burn recently.

 

[Alan Whitmore]

I wonder if we could do an experiment...

Choose one popular HPR reloadable motor (J maybe) and have various people buy them through their local dealer, build them and deliver them at a national launch. Then test them all and see what variation there is in practice for a "normal" motor. Then we'd have a basis for evaluating the variations in the N4000.

This would be a great experiment, I would love to see that data! The main problem under discussion is whether (and perhaps how) variation in the construction techniques of different users assembling motors that require that grains be glued in might cause significant variation in performance. I don't know of any 54mm motors that require grain gluing, so that would restrict us to 76 or 98mm motors, which cost A LOT. Most people aren't that eager to spend money for motors that would not be flying their rockets. Can't blame them.

I do have some data that many people have not seen that speaks to reproducibility. Immediately after I took over as TMT chair, I talked the BoD into buying 12 Aerotech I161W reload kits for a statistical test. There was some noise on the wavelength at that time about whether the position of a motor during firing would affect the burn time of the delay elements. My merry band of TMT'ers and I worked out a way to fire 4 each in different positions. There was no significant difference in delay times.

But, the graphs give us a good chance to appreciate how much variation we can expect from a group of "identical" commercial motors. As I recall, there were various numbers of 3 different day codes in this batch of reloads. The total impulse numbers were very close, well within the NFPA 1125 criteria, but you can see that there was considerable variation in the max thrust points.

You can also see that there was significant "ringing" in the test stand, this was before I figured out how to damp the structure to prevent ringing.

Alan