Just Jerry

Still looking for an image of one firing.

Just Jerry

NiallOswald

Anyone know why the longer motors burn for less time? Something to do with increased mass flow though the grains closer the nozzle?

What's the propellant mass of the K635? I know the 5-grain 54 classic just sneaks in under 1kg, which is significant for UK users as it's the cut-off point for the recently granted exemption.

Oh, and when's 'Blue Jack' due?

NiallOswald

Just looked on the CTI website - 1115g Quoted ISP is ~30s lower than for classic.

denverdoc

Just a guess w/o more info--the longer motors have bigger cores, decreasing web thickness and may run at somewhat higher pressures which increases burnrate???

NiallOswald

With the exception of the 6G XL, the propellant mass divided by the number of grains is identical across all the motors.

I'd forgotten that the L is a different grain geometry, and it's the one that shows the biggest deviation. The trend is quite slight and doesn't hold for the 3 grain J motor. I did notice something similar in the 38mm motor curves though.

Interestingly (well, to me anyway) the thrust curve shape of the classic motors changes with motor length, becoming more steeply regressive with more grains, and in the case of the 6G/6GXL motors showing what I think might be staggered burnout of the propellant grains.

I may be wrong, but I was under the impression that within a particular motor line, ProXX motors used the same grains for ease of manufacturing. The Pro38 classic motors I've seen certainly look that way. Perhaps Anthony can shed some light...

denverdoc

Niall,

Well that is interesting and a nice piece of detective work to note that the mass/grain is constant. The erosivity is likely greater as you note; per your suggestion, maybe Anthony can provide a real answer vs surmise. As an afterthought, it occurs to me that if the core is constant, and length of grain segment is as well, the geometry will progressively change with the number of segments. In other words initial Kn should increase which may be offset by nozzle specs, but given the greater relative surface area burning, may account for shorter burns and quicker burn down...? In this regard some of the Loki motors are interesting using very short grains in smallish motors. The I316 comes to mind, but can't find a T-t curve for it. Very progressive IIRC.

Just Jerry

You will notice the 2G, 3G, 4G and 5G curves are of a similar shape but simply higher magnitudes. The 3G seems to be operating at a lower Kn possibly due to drill size coarseness.

The 6G and 6XLG has a step so there is a larger core near the nozzle.

It appears the 6 longer G also has bigger cores the entire length thus shortening the burn time.

The other explanations and theories I have read so far do not fit.

Just Jerry

jsdemar

FWIW, I agree with Jerry's assessment.

-John D.

denverdoc

well i would take exception to this--the relative area will vary as a function of number of grain segments. Here is why:

let N=number of segs; assume the top is inhibited, so 2N-1 is the number of discs burning,

The volume of the propellant is given by N*K where k is the volume of one segment.

The rtio then of volume to surface area initially is given by

N*k/((2N-1)*Ar+(N*Ac)) where Ar and Ac are the area of the intersegment discs and core area respectively. assume that L is 1.5 times the diameter of the grain and the core is 1/3 the diameter.

Then the SA of the disc is Pi*8/36*diam^2 and the sa of the core is Pi*1.5diam^2.

The cores contribute more by a factor of 6, so it can be expressed as
N/((2n-1)Ar+6N*Ar)) whwn N=1, 1/(1+6) when N=3 3/(5+18) and when N=10
10/(19+60)

In other words a subtle effect which varies between 1/7 and in the limit as N approaches infinity, 1/8. So maybe a 20 percent difference in the extreme. as burn progresses the effect is amplified.?

Hey this is all conjecture, but nothing better to do at the instant.....

denverdoc

oops blew the math there, the core surface area should be Pi*1.5/3*dia^2 and so the impact on volume of fuel/surface area lit is larger than what I suggested

ddmobley

Still cuddling my DVM until Doc rechecks resistance math.

denverdoc

LMAO. hey algebra (and punctuation/grammar/spelling for that matter) is not for the attentionally impaired chaps like your's truly. Besides I only missed by a factor of 3, so where's the problem?

Just Jerry

One would expect a momentary difference in curve shape at ignition, transitioning to no effect in under 1/4 of 1 second due to erosivity on motors with high grain counts, which also have virtual chokes at each segment joint due to lateral flow, but to some degree erosivity is a binary condition, happening or not.

I had one motor once that expended 20% of its total impulse in the ignition erosivity spike WITHOUT failing (try doing that twice!), and these motors seem to exhibit more of a pellet charge than an erosivity spike. The transients onthe largest motor could be either data noise or pogo.

Just Jerry

denverdoc

Jerry,

Well all I can see is that the time to the start of cutout begins progressively sooner as N goes up, which would follow from a greater SA/Vol ratio. Why this headstart would wash out as you suggest is beyond my meager understanding of propellant burn dynmics.

The shelving Naill mentions is also present. But what about the notch seen in the shorter motors that progressively disappears--what accounts for that?

NiallOswald

Doc - I think I see what you're saying (equations without a proper editor are never fun to decode!) but AFAIK the grains are all identical and all burn on both ends and the core. Certainly the case for Pro38 motors.

If Anthony turns up - I have another question. How come (as far as I have seen) the VMax reloads don't use the BP grains for ignition? Core too big?

NiallOswald

Jerry - on the subject of test stand artifacts, have you seen the plot for the AT H999N? Needs more damping, badly: http://www.aerotech-rocket..._Sheets/h999n-p_info.pdf

J1999N is pretty bad too.

NiallOswald

3rd post in a row, sorry. I just knocked up a spreadsheet to try and replicate this ringing and see what effect it would have on the total impulse measurement. I compared an 'ideal' motor represented by an 1100N step at t=0 to the same step with ringing of similar magnitude and time constant to that on the H999 plot. Over 0.2s (by which point the ringing is negligible) the net result is that the trace without ringing integrates to 220Ns (as expected) whilst the one with ringing gives 222Ns. Less than 1% error, not really worth worrying about in the grand scheme of things.

I've attached the file if anyone's interested. Feel free to poke holes in my maths

UncleVanya

I find it interesting that nearly the entire difference is accounted for by cell 13 (12th data point).

denverdoc

Niall,
Being a fan of that motor, I was curious about the ringing effect as well. (I don't think that is what causes the notch in the CTI data as Anthony is scrupulous re damping of his test apparatus and it disappears besides with bigger initial kick in case this was implied). I'm glad to see that you did the fit and confirm it was a wash. I'll use a math editor next time Only wish it could do the algebra--this is for formatting only unfortunately.

denverdoc

that inhibited top segment makes all the difference, wonder why I assumed it was inhibited. Just having a bad hair/brain day I reckon.

NiallOswald

Yep, it's pretty much all in the first couple of overshoots. From eyeballing it, it looked like the area 'above' would be greater than the area 'below', I wanted to check.

Of course my 'ideal' profile is a massive simplification, in reality is there likely to be a large (but brief) initial 'spike' which could set up the ringing, or is it just the rate of rise of thrust causing it?

My suspicion is that this will occur most strongly with small, high thrust motors with a rapid initial rise of thrust. From my basic understanding of solid propellant characteristics, I understand that Warp9 has a fairly low pressure exponent value (0.287) and a very high rate coefficient (0.123) (from http://www.specificimpulse.com/images/ISP_Portfolio.pdf) which would suggest rapid start-up. Compared to most motors of the same thrust, it's much smaller - now if I can just get my head round the L-C-R equivalent circuit of the test stand mass-spring-damper system I'll be able to back that up with some sums.

Something else I have also been pondering of late is hybrid start-up, which often generates large transients on the thrust curves (sometimes quoted as the motor's maximum thrust - incorrectly I feel). It occurred to me that with many HPR hybrid designs, at the moment when nitrous flow commences, nitrous will actually be entering the chamber from two injector ports - the bulkhead injector and the fill tube/stem. I certainly know from watching videos that on monotubes the fill hose continues to vent visibly after leaving the motor.

At the moment of bursting, with low chamber pressure, I can imagine that a significant flow of nitrous into the chamber from the fill hose occurs, perhaps explaining the transient spikes seen? Given that the nozzle is also partially blocked at this point I was left wondering how they don't CATO every time! I expect there's probably some degree of negative feedback as the chamber pressure builds - any ideas?

denverdoc

high freq and high amplitude--the extreme example of which is the Delta (dirac) fx: infinite amplitude, and infinitesimal duration, the area of which=1, cousin to whom is the step fx.

That will cause a system to ring in such a way to completely characterize it (impulse response, transfer fx). (Also a slick way to handle those second and higher order systems math wise when you convert to freq domain using LaPlace or Fourier transforms.)

I think of those W9 motors as a real life step function....Only flown one, but however brief, most memorable. Truly a scalded cat/bat out of hell motor if there ever was one.

NiallOswald

I've flown the CTI I800 which while not quite in the same league as Warp9, made for a pretty punchy 40G liftoff. Someone said 'was it supposed to do that' Pretty big flame for an I motor too: http://www.flickr.com/phot...wald/2524222884/sizes/l/

Anthony Cesaroni

I assume you mean the ferrocene league.

Anthony J. Cesaroni
President/CEO
Cesaroni Technology/Cesaroni Aerospace
http://www.cesaronitech.com/
(941) 360-3100 x101 Sarasota
(905) 887-2370 x222 Toronto

Just Jerry

Anthony makes a good point. He achieves his result without resorting to ferrocene compounds, which is a bit of a technical feat.

I suspect Gary is not subject to much scrutiny on that point or we would not have access to Warp-9.

Just Jerry

NiallOswald

I like the way your VMax press release makes a point of not using burn rate catalysts whilst Aerotech's site brags about being 'not afraid' to do exactly that. Excellent marketing on the part of both companies Will there be longer 38mm VMax motors ever?

I've always thought that using a catalyst called CATOcene is asking for trouble...I have heard 'interesting' stories about warp 9 demo flights at LDRS 17.

As it happened, the delay snuffed on my I800, but it was a non-issue thanks to electronic ejection.

Anthony Cesaroni

Are you sure the delay snuffed? How much of the delay burned? Any photos would be helpful.

Anthony J. Cesaroni
President/CEO
Cesaroni Technology/Cesaroni Aerospace
http://www.cesaronitech.com/
(941) 360-3100 x101 Sarasota
(905) 887-2370 x222 Toronto

NiallOswald

Well, it either snuffed or was never lit in the first place - what is the likelihood of that occuring?

I'll get a photo this weekend, either way the vast majority of the delay element remains. The accelerometer data is consistent with the published thrust curve.

Anthony Cesaroni

Those delays are very transient tolerant so the latter I suspect possibly due to igniter placement. Please contact me off list with any information you may have.

Anthony J. Cesaroni
President/CEO
Cesaroni Technology/Cesaroni Aerospace
http://www.cesaronitech.com/
(941) 360-3100 x101 Sarasota
(905) 887-2370 x222 Toronto

jsdemar

You may also notice that the three different grain counts for that series of Warp 9 motors have identically-shaped curves. In other words, they were apparently just scaled to match the total impulse and are not the actual thrust curves.

-John D.

Just Jerry

It's easier that way. He has a Macintosh!

The CTI method of publishing FULL data with reference to file names, is so weird I think I need a drink. Okay, two.

Just Jerry

Just Jerry

Watch out. Like Bentley, he might send a full team of analysts to your site to analyze the issue to assure it does not propagate across the civilian product line.

Don't give him your address or GPS coordinates!

Just Jerry

denverdoc

Not only that but best I can figure all the 999's in the various designations are more about marketing and no longer indicate the actual average thrust of the motor.

when you have that much ringing, could br anything between 970 on up, depending on how you massage the data.... Even wondered if Gary detuned his rig to act like some spring loaded cell just for the occasion.....

NiallOswald

How should I go about doing this, since I do not appear to be able to contact you via PM on here. I have a couple of photos of the offending delay to send you.

Thanks.

Garoq

Actually the H999N curve is not from our stand. For comparison, the I1299N curve was obtained from AeroTech/RCS equipment.

http://aerotech-rocketry.c...m/38_480n_in_20069-1.pdf

denverdoc

That looks very clean indeed. Just giving AT some grief re the 9's that appear in all the W9 motors--no one is that good, and I really couldn't care less so long as w/in a few percent.

For the record, I think of it as creative marketing, and am not put off by whatever additives make em so fast.

PS: Using the data posted, ie 430N produced in 0.34 seconds I get an I-1264 so no one can accuse you of inflating the numbers

Just Jerry

The H999 curve with the waves in it is sufficiently "special" you should go into great technical detail as to how that came about, either for artistic purposes or as a teaching aid regarding test stand use.

BTW which stand and user(s) generated that curve anyway?

Just Jerry