Friday, July 31, 2015

A drag of a topic...

As I mentioned in previous posts, this year's TARC competition will probably see many teams struggling to loft their two eggs to altitude without busting the weight limit on the rocket. Most teams use the Tim Allen philosophy when it comes to increasing altitude - use a bigger motor (more power!). But there is another way, one that can result in a gain of tens of feet in the rocket peak altitude (maybe as much as a hundred). This approach, used by contest rocketeers all over the world, is largely ignored by TARC teams because it adds considerable work and time in building the rockets. Drag reduction can't be done in 30 minutes, but the math and years of experience shows that it can produce significant improvement if done properly.

Long before we had personal computers, a lot of smart folks were looking at drag and its effects on rocket flight. The math is fairly complicated - at a basic college calculus level - but even back in the 60's and 70's, advanced rocketeers were producing tables and formulae to estimate the amount of drag a model rocket would experience in flight. A oversimplification would characterize the drag by a single dimensionless number, Cd, called the drag coefficient. A flat surface has a drag coefficient well above 1, whereas a sphere has a Cd around 0.45. Most of my rockets seem to have drag coefficients around 0.75 - at least that's what I get when I try to match the simulations to the altitude profile given by the altimeters. The deal with drag is that you want to keep the airflow smooth around the rocket - any disruption, whether it is caused by rail buttons, base drag at the rear of the model, rough finish, or improper match of the nose cone to the body tube, increases the drag coefficient and reduces the altitude. TARC rockets need rail buttons or launch lugs (unless you want to spend the bucks to build a launch tower), so not much can be done there. However, every rocket can have a smooth finish - all that takes is additional work.

Rocksim screen capture showing Cd override panel (Click to enlarge).
To get an idea of how altitude depends on drag coefficient, I fired up my copy of the Rocksim software and loaded a past TARC design, the Over Easy. Rocksim has several advantages over Open Rocket in the technical area - you can design models with external pods and fins on fins, you can run simulations to figure out the optimum rocket weight, and you can put in your own values of the drag coefficient for the model (Open Rocket lets you do something similar by allowing you to choose the component finish - rough, unfinished, regular paint, smooth paint, and polished - but you just can't put in a number). I kept everything constant (weight, motor, etc)  except for the drag coefficient, which I varied from 0.55 to 1.05. The results are below - note that the altitudes range from 730 to over a thousand feet!

Effect of varying the Cd.
The catch is that a smooth finish requires paint, which adds weight to the rocket. So, as with anything in the real world, there is a tradeoff; you want as smooth a finish as possible without adding so much weight that you eliminate the altitude gained by reducing the drag. Experience has shown that a one or two primer coats, each followed by a vigorous sanding with increasingly fine grit sandpaper, and then a base coat of paint can yield a pretty smooth finish with only a modest weight increase.

Back in 1970, Estes produced Technical Report TR-11 "Aerodynamic Drag of Model Rockets". I very much recommend that all TARC teams and serious rocketeers read this, even though it is filled with math and stuff. At a very minimum, they should read pages 39-47, where there are some simple guidelines for drag reduction and flight test data for various configurations and finishes for an Estes Alpha. By airfoiling the fins, adding a boat tail, and putting a smooth finish on the model, they were able to improve the Alpha's measured altitudes from 319 to 446 feet, which is huge! This is a increase of about 30%, which is in line with the Rocksim simulations I performed for the Over Easy.

In case you can't locate a copy of TR-11 online (Google "Estes TR-11 drag"), here are the 5 rules for drag reduction:
From the beginning of model construction, take time to sand all parts for a good fit: match the nose cone - body tube junction carefully, round the leading edge and sharpen the trailing edge of the fins. This initial work is a great step toward the reduction of the pressure drag of the rocket.
Correct fin alignment will keep the model from rolling during the ascent; this will eliminate unnecessary induced drag caused by the fins twisting through the air at angle of attack. Misalignment of the launch lug can cause flow separation on the body tube and excessive pressure drag.
Besides giving a high quality appearance to the model, the smooth finish delays transition of the flow to the high drag turbulent boundary layer condition. Even when the flow is turbulent, a slick surface will have less drag than a rough surface; so to reduce skin friction drag get the model rocket finish mirror-smooth.
Reduce the interference drag by filling in the fin-body tube junction to guide the air smoothly past the fins.
Anytime the body tube diameter is greater than the engine diameter because of some special design feature, boat-tail the model. The base drag, which contributes an appreciable fraction of the total drag will be cut drastically by the boat-tail.

The moral of this post? A slick rocket is a good thing.