Streamers and the Elusive Nature of Truth

One reads in several places, including the 2004 edition of Stine and Stine’s Handbook of Model Rocketry and the 2008 edition of Van Milligan’s Model Rocket Design and Construction, claims along the lines of “Research done by Trip Barber and others at the MIT Rocket Society indicates there is an optimum streamer size for every model rocket design. Generally, the optimum streamer size to obtain the slowest descent rate has a length-to-width ratio of 10:1.” [Stine and Stine]

It ain’t so!

That research (T. Barber and T. Milkie, “Streamer Duration Optimization”, J. MIT Rocket Soc.) was done in 1972 and there’s been a good deal more done since then, and the more recent research indicates the “10:1” claim is just not correct — at least, not for all materials and all methods of preparing a streamer. Barber and Milkie themselves apparently showed 10:1 was optimum only for crepe paper, though they tested other materials. Further research was done by C. Flanigan in 1976, Charles M. Sykos in 1980, and B. Kaplow and A. Jones in 1984. The most recent research I’ve found is Chris Kidwell, “Streamer Duration Optimization: Material and Length-to-Width Ratio”, presented at NARAM-43 (2001). One of Kidwell’s results was that for folded 1.5 mil Mylar, 4″ wide, streamer performance improves as one goes from 20″ to 60″ long; no optimum length was found.

Kidwell also tested a number of other materials, though not with varying lengths. He makes claims about differing speeds due to differing materials, though to my eye the statistical significance of those results is pretty low. But un-folded crepe paper does seem to perform worse than folded streamers of any other material tested — so even if 10:1 is optimal for un-folded crepe paper, un-folded crepe paper itself isn’t optimal.

So as of 2001, and probably by 1984, there seems to have been good reason to dump or at least doubt the “rule” about the optimality of 10:1 streamers. Yet Stine and Stine in 2004, Van Milligan in 2008, and numerous online commenters up to 2013 continue to perpetuate it. Ain’t received wisdom wonderful?

 

Advertisements

3 thoughts on “Streamers and the Elusive Nature of Truth

  1. Van Milligan did some research around 1991, presented at a mini-convention at the North Georgia Regional Meet in May of that year. Though the presentation was mostly about folding techniques and fold shapes, I think his findings were that the 10:1 ratio still held the best results. The popular material for competition streamers at the time was MicaFilm, a model airplane covering from Coverite that I believe is no longer available. Ed LaCroix/Apogee was challenging this with his ‘DragStrip’ material, which was basically transparent red mylar. For his research, Tim was using architects’ tracing material, because it was easier to fold than MicaFilm or DragStrip.

    If I remember correctly, longer streamers did have more drag, but their added weight and difficulty in folding them compactly made them less desirable than a 10:1 streamer. There was also some idea about the flapping action of the streamer being damped as it got longer. I’ll have to go read Chris’ report, but since competition streamers remain mostly 10:1 to this day I don’t think it is merely because Stine said so.

    • Thanks for that information. On a quick Google search I didn’t come up with anything about that research specifically. But Peak of Flight issue 128 (dated 2004) has an article on streamers by Van Milligan. I’ll need to read it more carefully but on first look I don’t see anything that specifically says 10:1 gives optimal performance. What he does say is “A long while back, I did some digging into all the science fair reports that I could get my finders [sic] on. The result is the graph shown in Figure 2” — and said graph, of L/W vs. drag coefficient, has exactly three points on it. No error bars, and no mention of what material the data are for, or whether the streamers were folded or not, and so on. Not much to hang a thesis on, but taken at face value, it doesn’t show any optimal L/W point. As you say there may be good arguments in favor of 10:1 streamers, having to do with weight, compactness, and so on. But my point is that people go around claiming that 10:1 streamers give optimal performance, regardless of material or thickness or folding, and I don’t see that the data support such an idea at all.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s