Off the rail velocity POLL

[jqavins]

I disagree, Rail length is VERY important for exit velocity. A longer rail allows a longer time to accelerate while guided therefore obtaining a higher exit velocity.

I guess I was unclear. Yes, of course rail length is super important in developing exit speed. The longer the rail the better the speed, so a longer rail is needed when there is lower initial thrust.

The original question was what exit speed is needed (and the discussion veered into the question of T:W). The answer is most often considered to be (I gather from this thread) 45 fps, so lets go with that. If you have 45 fps and the end of a three foot rod, at the end of a six foot rail, or at the end of an eight foot rail, it's the speed that matters, not the length.

The OP asked about rail exit velocity, not how to get there. Rail length might determine that value for a given rocket/motor combo, but it doesn't matter to the OPs question. 45 fps off an 8ft rail is no different from 45 fps off a 6ft rail from a safety perspective.

Oops, ninjad.

Good to know. I had calibers stuck in my head based on this figure which Dahlquist published. He wasn't the MIT guy, that was Galejs and I honestly haven't read his report...BUSTED

The Galejs paper was a revelation for me. I spent years wondering why 1 caliber is some sort of magic number. I mean, anything over zero means the lift-induced torque is in the right direction, and of course you want something more than a tiny margin due to inevitable inaccuracies in calculations, but why 1? When a ratio of this to that is important, and you learn that 1 is an important figure, i.e. this equals that, there's usually some underlying reason. So I asked around. I asked here. I asked Tim the Rocket Man van Milligan. I asked some other places, and no one had an answer. Finally I ran across the Galejs paper and received enlightenment. I was so pleased that I wrote to Tim to suggest that he reprint the paper in Peak of Flight, and gave him Galejs's contact information; lo and behold, there it is, in issue #470.

And, ironically, there is no fundamental underlying reason, it's just a good empirical rule of thumb that happens to be 1.0.

I think we’re talking about the same thing. You need a greater static margin with long skinny rockets is because the greater moment of angular momentum reduces the ability of lift to quickly restore the course of a rocket.

I don't think we are talking about the same thing. You're not wrong about the moment of inertia, but I'm talking about the movement of the CP forward with increasing angle of attack. The effect of body lift increases, the CP moves forward, and it can move past the CG. And the movement is more or less proportional to the length (and other factors) not the diameter, so we need more diameters of margin when there's more length to assure the CP stays behind the CG, where it belongs.

 

[Tractionengines]

@Steve Shannon what is the thoughts about a dual thrust motor in the 5:1 / 3:1 TTW ratio spec on the codes... the codes all reference Average thrust, but a motor specifically with high initial thrust can get a heavier rockets to a safe speed quickly, (shorter rail) then drop to the sustain part of the thrust curve...

ie. The Aerotech K455NW-P

Is there any flexibility left for the RSO, LCO, and Flyer to agree on a "safe flight profile", for motors with "non-standard" thrust profiles.

 

[Steve Shannon]

@Steve Shannon what is the thoughts about a dual thrust motor in the 5:1 / 3:1 TTW ratio spec on the codes... the codes all reference Average thrust, but a motor specifically with high initial thrust can get a heavier rockets to a safe speed quickly, (shorter rail) then drop to the sustain part of the thrust curve...

ie. The Aerotech K455NW-P

Is there any flexibility left for the RSO, LCO, and Flyer to agree on a "safe flight profile", for motors with "non-standard" thrust profiles.
View attachment 541178

Great point. There’s definitely flexibility. Initial thrust can make all the difference. Truly the only important portion of the thrust curve with respect to velocity off the rail is the period before the rocket leaves the rail. That’s what they should focus on.

 

[Steve Shannon]

I guess I was unclear. Yes, of course rail length is super important in developing exit speed. The longer the rail the better the speed, so a longer rail is needed when there is lower initial thrust.

The original question was what exit speed is needed (and the discussion veered into the question of T:W). The answer is most often considered to be (I gather from this thread) 45 fps, so lets go with that. If you have 45 fps and the end of a three foot rod, at the end of a six foot rail, or at the end of an eight foot rail, it's the speed that matters, not the length.


Oops, ninjad.



The Galejs paper was a revelation for me. I spent years wondering why 1 caliber is some sort of magic number. I mean, anything over zero means the lift-induced torque is in the right direction, and of course you want something more than a tiny margin due to inevitable inaccuracies in calculations, but why 1? When a ratio of this to that is important, and you learn that 1 is an important figure, i.e. this equals that, there's usually some underlying reason. So I asked around. I asked here. I asked Tim the Rocket Man van Milligan. I asked some other places, and no one had an answer. Finally I ran across the Galejs paper and received enlightenment. I was so pleased that I wrote to Tim to suggest that he reprint the paper in Peak of Flight, and gave him Galejs's contact information; lo and behold, there it is, in issue #470.

And, ironically, there is no fundamental underlying reason, it's just a good empirical rule of thumb that happens to be 1.0.


I don't think we are talking about the same thing. You're not wrong about the moment of inertia, but I'm talking about the movement of the CP forward with increasing angle of attack. The effect of body lift increases, the CP moves forward, and it can move past the CG. And the movement is more or less proportional to the length (and other factors) not the diameter, so we need more diameters of margin when there's more length to assure the CP stays behind the CG, where it belongs.

You’re right; I should have called it moment of inertia. I agree also the the change in location of Cp wasn’t what I was talking about. But the greater moment of inertia is also important.

 

[Mugs914]

If the rail buttons aren't slightly meted it was too slow off the rail...

 

[jqavins]

Great point. There’s definitely flexibility. Initial thrust can make all the difference. Truly the only important portion of the thrust curve with respect to velocity off the rail is the period before the rocket leaves the rail. That’s what they should focus on.

I wonder, by way of 99% pointless musing, if there's a worthwhile way of quantifying this as a new sort of motor rating and associated safety rule (of thumb). Impulse provided in the first 100 ms, or 200 ms, for example. Catalog that for available motors and specify a safe minumum ratio of that to mass (or weight).

Oh, gee, wait. Since speed is momentum over mass, the quotient of "early impulse" over launch mass is the speed at the end of that early period. So, once a good period for the purpose is chosen, simply divide that catalogued figure by the mass and it's got to be at least 45 fps. Hmm...

You’re right; I should have called it moment of inertia.

I hadn't noticed that you didn't.

 

[rharshberger]

I wonder, by way of 99% pointless musing, if there's a worthwhile way of quantifying this as a new sort of motor rating and associated safety rule (of thumb). Impulse provided in the first 100 ms, or 200 ms, for example. Catalog that for available motors and specify a safe minumum ratio of that to mass (or weight).

Oh, gee, wait. Since speed is momentum over mass, the quotient of "early impulse" over launch mass is the speed at the end of that early period. So, once a good period for the purpose is chosen, simply divide that catalogued figure by the mass and it's got to be at least 45 fps. Hmm...

I hadn't noticed that you didn't.

Most RSO's afaik are concerned about the first .5 seconds of thrust, after that as long as the thrust greater than 1:1 (and the rocket is already stable) the rocket SHOULD continue to be stable.....theoretically....

 

[OverTheTop]

I wonder, by way of 99% pointless musing, if there's a worthwhile way of quantifying this as a new sort of motor rating and associated safety rule (of thumb). Impulse provided in the first 100 ms, or 200 ms, for example. Catalog that for available motors and specify a safe minumum ratio of that to mass (or weight).

It's called a simulation. If the rocket is suspected of being too heavy for the motor a simulation is good for calculating exit velocity and adding some certainty to the decision. It removes a lot of the guesswork. I had this at a recent launch where I was RSO. A flyer only had a short rail available and he was able to run a simulation of his rocket and confirm rail exit velocity was suitable. Flight was nominal.

 

[Tractionengines]

Most RSO's afaik are concerned about the first .5 seconds of thrust, after that as long as the thrust greater than 1:1 (and the rocket is already stable) the rocket SHOULD continue to be stable.....theoretically....

I AGREE, and I have had this discussion with others. What you and Steve just said is what I feel should be in wording of the codes. But every code I think says "average thrust". Maybe it should be recommend that the wording of the TTW ratio of the codes be reviewed. As more grain geometries and mixed propellant motors come into mainstream use, the main issue is to get initial stability.

Here is where I agree having a SIM shows what should be expected of the flight. [If the SIM is correct.]

Mike

 

[jqavins]

It's called a simulation. If the rocket is suspected of being too heavy for the motor a simulation is good for calculating exit velocity and adding some certainty to the decision. It removes a lot of the guesswork. I had this at a recent launch where I was RSO. A flyer only had a short rail available and he was able to run a simulation of his rocket and confirm rail exit velocity was suitable. Flight was nominal.

Of course, a good quality sim is the best. Without having that for all the rockets that come to the RSO table, it's necessary to have good rules of thumb. So, I muse about the possibility of a better one, i.e. one that's more helpful. But yes, just musing.

 

[Steve Shannon]

Here’s what the Tripoli Unified Safety Code says about T/W:
“The thrust-to-weight ratio of a rocket typically should be at least 5:1. However, the RSO may approve a thrust-to-weight as low as 3:1 ratio. Initial thrust-to-weight ratios lower than 3:1 may only be authorized by an RSO if an active stability system is included.”

 

[rharshberger]

Here’s what the Tripoli Unified Safety Code says about T/W:
“The thrust-to-weight ratio of a rocket typically should be at least 5:1. However, the RSO may approve a thrust-to-weight as low as 3:1 ratio. Initial thrust-to-weight ratios lower than 3:1 may only be authorized by an RSO if an active stability system is included.”

no specification of average or max thrust, the way it should be.

 

[Richard Dierking]

This is all based on calculated exit velocity. Keep in mind that as soon as the rocket is over 1:1 it's going to start to move and the exit velocity is different than calculated/simulation. Personally, I think that our rails should be longer. Most of the time, the length is just what is available; like you cut a 12' in half and you have two 6' rails, or a 20' in half and you have the two 10's. And, the 6' go on the middle row and the 10' go in the back. Sometimes you end up with a 12' and a 8' for the 1515's.

For many years, I have wondered why there is so much dispersion at institutional-types of launches. We need to study and understand this better and many people building and flying their rockets need to spend more time understanding what their rockets are actually doing and making necessary changes.

I would say, sim at 45 fps exit, actually measure it at launch with observations, and make appropriate changes.

 

[Zeta]

Does prox·im·i·ty to the blast plate, or the impingment angle of the plate play a roll in initial lift off?

 

[jqavins]

This is all based on calculated exit velocity. Keep in mind that as soon as the rocket is over 1:1 it's going to start to move and the exit velocity is different than calculated/simulation.

????

The sims use [thrust minus weight minus drag] as the force on the rocket. You seem to be saying that once the rocket begins moving, that's not valid. Am I wrong, did you mean something else? Of course, the simulation results are never perfect due to small errors in all those forces and the rocket's mass, but it seems like you mean something else.

Does prox·im·i·ty to the blast plate, or the impingment angle of the plate play a roll in initial lift off?

Without experimental or theoretical proof, I am comfortably certain all the same that effect of blast impingement is somewhere between trivial and non-existent. Except in a tube launcher.

 

[Tractionengines]

Does prox·im·i·ty to the blast plate, or the impingment angle of the plate play a roll in initial lift off?

Only "Rod whip" ... if an angled plate is attached to rod/pad it pushes the rod / pad / rail, etc. Having the blast deflector assembly separate from the launch pad assembly can allow for straighter launches.

 

[Richard Dierking]

We can see from observations and replay of video that a slanted plate blast deflector can jog the bottom of the pad and therefore affect the rail or rod. This can affect departure angle, and the snap could also slow the velocity. Grain geometry could also impart some additional friction of the rail buttons on the rail. How significant, I don't know.
If you wish, I did some work on Rail Exit Velocity - Simulated vs Actual and you could read about that. I might continue this work because of my roll control project. I'll be using the same rocket and motor for about 5 flights and the analysis of video at 240 fps seems to work well for determining exit velocity.

So, I wasn't clear about the 1:1 thing, sorry about that. What I meant is that you could sim for a 6' long rail but the motor is pressurizing up and the rocket begins to move. You know, that sometimes short increase in the thrust curve after ignition. And, it's not just the motor, it's the igniter stuff too.

Again, there would have to be a lot more work to say anything definitive on simulated vs actual exit velocity. The only reason I brought this up is that the conversation was about how high the velocity would have to be at exit to be stable/safe, and you might have to provide some increase in the simulated to meet the actual minimum requirement. Hope that makes sense.

 

[waltr]

Here is another paper that directly answers the OP's question:
https://www.apogeerockets.com/Wind_Caused_Instability
Also note that Open Rocket does calculate the change in CP with change of the Angle of Attack.

 

[Tractionengines]

A few more notes on this topic...

1) Always assume guiding stops as soon as you no longer have 2 points of contact. Rail buttons do not stop any rotation, so as soon as there is only one button; the rocket on free to rotate around that button while its sliding up the rail. [Yes. If you have rail guides, or a tube around a rod, the lowest point will give you some guidance. BUT if the stability is not sufficient when the point above that clears the rod or rail; you get more torque applied at the aft point and it can break loose. ]

2) Make sure to measure USABLE ROD/RAIL LENGTH. Some people talk about overall rod length. They buy one "x" feet long and put that in the SIM. Then they build the pad and use "a" amount of length to mount it to the structure. Then put a blast deflector "b" distance above that. Then put a stop pin "c" distance above the blast deflector for the aft end if the rocket to sit on. THEN THEY ASK WHY IT ANGLED WHEN LAUNCHED, or blame it on weathercocking, or poor motor startup, etc. (Most RSOs will catch this just from experience, but new Rocketeers may over look it.)

So an extreme example: If you have a rocket with the aft button 6" up from the end of airframe. The forward button 18" in front of that. You bought a 6' (72") rail so it fits in the car. Then built your pad around a "Jaw Stand". So you clamp 12" in the jaws, your blast plate is 6" up from that, your stop pin is 6" up from that. The actual GUIDED length is ONLY 24". That 24" is what should be used in the SIM.


(Note that since this is in the HPR Forum this answer does not apply to LPR where a single longer tube us used on an 1/8" rod for small rockets.)

 

[jqavins]

That suggests one should (among other things) design the launch stand to minimize the "lost" length. For example and for starters:

• Insert and secure a close fitting rod at least 6" into the center hole on the bottom end of the rail, with at least 6" of rod extending out, and use that rod in the jack stand.
• Also make the blast deflector with a hole that takes the same rod, so it site between the rail and stand.

You just got 18" back, giving you 42" instead of 24", but which is still a far cry from 72".

(My home club's rails are set up this way, and I assume many others are as well.)

 

[Alan15578]

The number in the motor designator, G80 for example, is the average Thrust in Newtons.
To convert Newton force to kilogram (kg) force divide by 9.81 (Earth's gravity).
For easy calculation divide by 10 (I can do this in my hear by moving the decimal point). This also gives a slight margin.
This gives a rocket weight in kg for 1:1 TTW.
For 3:1 divide by 3.
Therefore, For a G80, divide by 10 is 8, divide by 3 is 2.6kg rocket weight. or 80/30.

Note: At many launches the RSO will require a 5:1 Average thrust to weight. So the maximum weight of a rocket on a G80 is 80/50 = 1.6kg.

The reverse calculation is starting with rocket weight in kg.
Multiple the rocket weight in kg by 30 (for 3:1 TTW) or by 50 (5:1 TTW) to find the minimum average motor thrust.
So a 2kg rocket requires a 60N motor for 3:1 or a 100N motor for 5:1.

Just a minor nit. G80 was a poor choice. The G80 I am familiar with is a MR motor that is essentially a G100 with a measurable tail off thrust that lowers the average thrust to the 80N limit required for MR motors. In spite of the codified Average T/W ratio, Steve Shannon indicated farther down this tread that for some motors the initial trust should used. I'm waiting to see if Steve walks that back or clarifies his post, or if he intends to get the codification changed.

 

[Steve Shannon]

Just a minor nit. G80 was a poor choice. The G80 I am familiar with is a MR motor that is essentially a G100 with a measurable tail off thrust that lowers the average thrust to the 80N limit required for MR motors. In spite of the codified Average T/W ratio, Steve Shannon indicated farther down this tread that for some motors the initial trust should used. I'm waiting to see if Steve walks that back or clarifies his post, or if he intends to get the codification changed.

I’m not sure what’s not clear or what I would walk back. And I have no intention to suggest a change to NFPA 1122 and 1127.

 

[John Kemker]

I voted frigg'n ripp'n. I like Loki and used to fly a lot of Kosdon, so go figure. [grin]

Though, I am REALLY wanting to try an I110 in 38mm...

 

[Zeta]

finally!

 

[Richard Dierking]

So, now as a follow-up someone needs to build a rocket meeting the OP criteria launch it using different motors with a sim of 35, 45, and maybe 60 fps exit. Then, determine the actual exit velocity and describe the launch. A Feather Weight tracker would be best for the data, and show the trajectory path on Google Earth.

 

[jqavins]

I looked into video cameras with high frame rates yesterday. There are phone cameras that do 240 frames per second (FPS) and anything beyond that seems to be well over 1000 FPS, and several thousand dollars; there doesn't seem to be anything in between Or, a good camera (1500 FPS at 1080p, 3200 FPS at 720p, or higher rate at lower resolution) can be rented for a mere $450/day or $1800/week. Ouch! A group might manage to get such a camera for a day, with a large number of rockets ready to go. Still, even for a group pitching in, that's not cheap.

 

[Steve Shannon]

Maybe the easiest would be wind tunnel testing.
But really, what is the goal here? To find out the slowest off the rail velocity so that there is no margin of safety?
Our CP is estimated and could be incorrect; we need a margin.
Our knowledge of wind conditions higher than we can reach is estimated based on previous flights.
Our igniters and motors have tolerance bands surrounding their performance.
All these things can stack up one way to make a 45 fps rocket just barely stable or stack the other way to make a 30 fps rocket fly like a dream.
There are just too many variables to publish a number and say it’s golden.

 

[jqavins]

The topic seems to have drifted, caught in a cross-breeze. My last comment was about Richard's proposal for a comparison study of predicted and actual rail exit speeds, and the use of a feather weight tracker to determine the actuals. Previous work - was that you, richard - used video, and those images left considerable room for improvement; high speed video would be the best way.

I'll stop contributing to the threadjack.

 

[Richard Dierking]

I don't know how deep anyone wants to pursue this, but my work on measuring exit velocity started with a pretty simple question; Is the exit velocity given in programs such as Rocksim similar to the actual. It just seemed to me that some rockets looked slow coming out of the gate. If rockets are not reaching the anticipated exit velocity, then perhaps this could account for some of the dispersion I was noticing.

I looked into video cameras with high frame rates yesterday. There are phone cameras that do 240 frames per second (FPS) and anything beyond that seems to be well over 1000 FPS, and several thousand dollars

I tried a couple of things including a break-beam sensor at the top of the rail that would detect the two rail buttons passing by. Know the distance between the buttons and elapsed time and there you go! The break-beam was a bit unreliable because of environmental interference (the Sun mostly). But, I didn't pursue the laser break-beam sensor. This method would be the best, because the time in ms is displayed immediately after launch and you could measure any rocket that had a couple of rail buttons and you knew the distance between the buttons. Anyway, I settled on video analysis.

A GoPro at 240 fps works well. I mounted it on an extension poll. I used Aiseesoft Total Video Converter and Tracker for the frame analysis. Here's a couple frames from one of the test launches. Frames 1862 and 1865.

 

[jqavins]

I do remember that project, and I was part of rather extensive discussions. The pictures at 240 FPS are better than I remembered. Still, there's a good bit of motion between frames, and I'd love to see it done with a higher rate, 1000 or so, if only it didn't cost so darn much. Or the laser break beam, which I misremembered as having failed. Or the pair of microswitches with known spacing, which I also recall also failing.

Oh, wait, I said I'd stop threadjacking.