Mark Sullivan's Altitude Predictor ?

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

aerostadt

Well-Known Member
TRF Supporter
Joined
Oct 26, 2009
Messages
4,471
Reaction score
1,119
Location
Brigham City, UT
I am looking for Mark Sullivan's altitude predictor on the internet. My old website address does not work. Is anyone familiar with this old website and how it might be found?
 
I've been looking for it for nearly a month with no success. No search engines have provide any information except for the old domain appears to have expired.
 
I have it as a favorite in my computer, but nothing there when I click the link.
 
I wanted to use the predictor for my 4x OT. Seeing that Mark Sullivan's predictor was not going to appear anytime soon, I improvised a simulation on Rocsim 9. Surprisingly, the Rocsim prediction came very close to the old Sullivan predictor. My original estimate of drag coefficient in the Sullivan predictor worked pretty good.
 
I have 25 students building custom rockets. I fully expected to find markworld's altitude predictor (and delay time computer) online and functioning.

Anyone have any info about it?
 
What is special about the Sullivan predictor? Web-based simulators usually use simplified forms of the differential equations governing flight trajectory, like using constant thrust, constant air density, constant Cd. Thrustcurve is similar and will get you in the ball park for altitude and delay time. RockSim, OpenRocket, RASAero, and wRASP use numerical integration with fewer assumptions and are far superior tools.
 
What is special about the Sullivan predictor? Web-based simulators usually use simplified forms of the differential equations governing flight trajectory, like using constant thrust, constant air density, constant Cd. Thrustcurve is similar and will get you in the ball park for altitude and delay time. RockSim, OpenRocket, RASAero, and wRASP use numerical integration with fewer assumptions and are far superior tools.

In a word: simplicity. Essentially, enter body dia, motor, and CD and it spit out a fairly accurate flight prediction. True, there are way more advanced and involved methods, but that's not the intent. I really miss the Markworld program, if I was looking for ballpark estimates in speedy fashion, that was the program. As for accuracy, it depended upon the flight envelope; for most non-extreme mid-high power flights, I found it to be much more accurate RockSim, especially once the CD of a particular airframe was dialed in.

Give me three minutes with a rocket and the Markworld program, and I'll give you a more accurate speed and altitude prediction than the "superior" RockSim.

-Eric-
 
especially once the CD of a particular airframe was dialed in.

Yes, of course. Cd is critical to any type of simulation. Adjusting a constant Cd to match your altimeter's barometric altitude is somewhat helpful, but not robust. It becomes a catch-all for everything wrong in your flight. Wind tunnel, ballistic measurement, or CFD is the best way to get Cd vs. speed.

If Markworld (never heard of it) and Sullivan do what I think they do, then it is mathematically impossible for them to be "more accurate" than RockSim, OpenRocket, RASAero, and wRASP. The latter softwares capture the physics more precisely, assuming they are coded correctly, of course.

Ask yourself how "accurate" a simulation can be, and how do you confirm it with any statistical significance? Motor thrust and wind variation at the time of the actual flight are impossible to predict. If your electronics and simulations repeatedly correlate within 10%, then consider yourself lucky, call it good, and move on. Once dialed in, as you say, then RockSim, OpenRocket, RASAero, and wRASP should give you better odds than simplified equations.

The motor guide in Thrustcurve.org probably does what you are wanting, and gives you many motor results at the same time. Very handy. Perhaps a good compromise for you is wRASP, where you only need mass, diameter, and Cd, but the calculations are more precise. RockSim, OpenRocket, RASAero all require you to input the complete geometry of the rocket before you start simulating, and that is time consuming.
 
Yes, of course. Cd is critical to any type of simulation. Adjusting a constant Cd to match your altimeter's barometric altitude is somewhat helpful, but not robust. It becomes a catch-all for everything wrong in your flight. Wind tunnel, ballistic measurement, or CFD is the best way to get Cd vs. speed.

If Markworld (never heard of it) and Sullivan do what I think they do, then it is mathematically impossible for them to be "more accurate" than RockSim, OpenRocket, RASAero, and wRASP. The latter softwares capture the physics more precisely, assuming they are coded correctly, of course.

Ask yourself how "accurate" a simulation can be, and how do you confirm it with any statistical significance? Motor thrust and wind variation at the time of the actual flight are impossible to predict. If your electronics and simulations repeatedly correlate within 10%, then consider yourself lucky, call it good, and move on. Once dialed in, as you say, then RockSim, OpenRocket, RASAero, and wRASP should give you better odds than simplified equations.

The motor guide in Thrustcurve.org probably does what you are wanting, and gives you many motor results at the same time. Very handy. Perhaps a good compromise for you is wRASP, where you only need mass, diameter, and Cd, but the calculations are more precise. RockSim, OpenRocket, RASAero all require you to input the complete geometry of the rocket before you start simulating, and that is time consuming.

Buckeye, I'm not looking for an argument...I know I am in the minority here and do things a little differently than others: I'll say it: I don't particularly care for Rocksim for a number of reasons. I will always admit that RockSim is a very useful tool if used properly. I could expand on that, but I don't want to be labeled as a "hater"; I am not, I just have issues with a few of the program's faults. I do own a copy.

I have used all the programs you mention above; I like some more than others. In general, RASAero, Barrowman, and CAD are my methods if I'm working on a high performance or complex rocket. HOWEVER, for over a decade of high power flying I got by just fine simply by knowing a few basic parameters and using Markworld as a simulator. Though a subjective statement, I can say with confidence that using that simple altitude predictor was indeed more accurate than the (generally) overly optimistic world that RSim operates in. I have also submitted Class 3 paperwork, and I can tell you with certainty that RockSim is not the program that I'd choose to accurately define "extreme" flight profiles.

We can be thankful that there are many "methods of madness" available out there. The fact remains: for simple, basic, mid-high power flights, Markworld (Mark Sullivan's) Altitude Predictor was a quick tool that I really miss.

Your mileage may vary, and my methods are not necessarily the best for others.:cool:

Edit: Hadn't tried out the Thrustcurve motor guide in a while; some of the inputs are similar and the outputs are similar, but lacks things like launch angle and site elevation; as well as the velocity/acceleration/altitude: time graphs that were able to be manipulated.

-Eric-
 
Last edited:
In a word: simplicity. Essentially, enter body dia, motor, and CD and it spit out a fairly accurate flight prediction. True, there are way more advanced and involved methods, but that's not the intent. I really miss the Markworld program, if I was looking for ballpark estimates in speedy fashion, that was the program. As for accuracy, it depended upon the flight envelope; for most non-extreme mid-high power flights, I found it to be much more accurate RockSim, especially once the CD of a particular airframe was dialed in.

Give me three minutes with a rocket and the Markworld program, and I'll give you a more accurate speed and altitude prediction than the "superior" RockSim.

-Eric-
Quick, fast and dirty what-if calculations.

In addition to what Eric said, you could input a custom motor immediately. Say you wanted a 100 NS G1000 with an ISP of 2 NS/g. You input thrust = 1000 N, burn time = 0.1 s, and propellant mass = 50 g. Specify the loaded rocket weight, the diameter, and the CD and run the sim.

You could do clusters, multistage and cluster-multistage rockets just as quickly.

It was my goto jellybean sim for many years. Sorely missed.

Bob
 
Not sure if this will do, but try and search for Winroc. I have used that in the past, and it sounds very similar to what you're asking for. It's a free download.

Phil L.
 
Back
Top