Regarding the 30 degree spread making the lights harder to see.

Your eye is only going to capture a fraction of that 30 degrees, the rest is focused elsewhere, so it will appear more dim than if it were a narrower beam.

I had a choice with these LEDs (the brightest I could find) of (1) unfocused 90° FWHM, (2) 30° focused, or (3) 15° focused. I chose (2) as a reasonable compromise between (a) concentrating the light where I need to see it (down at the ground) and (b) spotting it somewhat off-axis as the rocket sways under the chute.

I work for an aerospace company, and I'm surrounded by pilots. My current project is doing flight tests with a Cessna 337, maybe I can have them take along a test strobe on our next flight test. Interesting idea. I do have a spare LED and MOSFET....

 

re 30° angle:
That will certainly be much better than the original 90° or wider pattern, but my thoughts were that in this case, narrower is better. I would've opted for the 15° beam, particularly considering the vertical distance you'll be attempting to observe the rocket from.

With a 30° spread, from an altitude of roughly 7,000 feet, the strobe will be illuminating a circle of roughly 1,800 feet, or over 1/3 of a mile, with an area well over a square mile!
With a 15° spread, same altitude, the circle would be roughly 900 feet for an area of 2,800 square feet.
[eta]
After deployment of the drogue chute, and later the main at 1,000 AGL, there will be nutation (swaying in a somewhat circular motion) of the rocket body centered underneath the 'chute for a period of time; this will make the rockets' strobe both more and less observeable (depending upon the angle of nutation) from a wider area on the ground.

If one of your test pilots could take a test strobe/beacon aloft, that could be very helpful - but don't hold your breath. Even something seemingly simple as opening a window and flashing a beacon out of it raises many cautionary flags.

 

I work for an aerospace company, and I'm surrounded by pilots. My current project is doing flight tests with a Cessna 337, maybe I can have them take along a test strobe on our next flight test. Interesting idea. I do have a spare LED and MOSFET....

Excellent!

Seriously, that would be useful information for many hobbyists to know, especially with regards to the OP, and similar ideas that aren't easily tested. Different lensed LEDs at 5k/10k/more feet depending on pilot mood and local congestion would be interesting photo compairson.

There are very intense LEDs out now, and getting better every couple of months. If a high output LED were to be pulsed near the manufacturer's suggested maximum overcurrent/cooling/rate, focused to 10 degrees in ACL installation, the LED might actually come out ahead in some ways. Much less EMI to shield from avionics, easy control from lower operating voltage, overall power use might be higher, but there are many benefits.

Lifetime/phosphor loss is another consideration, but that's for down the road. Xenon also have relatively short lives and diminishing light before failure (usually).

 

If one of your test pilots could take a test strobe/beacon aloft, that could be very helpful - but don't hold your breath. Even something seemingly simple as opening a window and flashing a beacon out of it raises many cautionary flags.

I've already talked to a couple of the guys, and they're open to the idea. But we're talking about through a window, not opening one and sticking the light out.

One guy did offer to attach the light to his R/C helicopter, but I doubt he'd be able to get it up to 5,000 feet.

 

italo,
I didn't even bother to search before I made my initial post. However, after searching:
http://www.google.com/search?hl=en&q=Anticollision+light+FAR
I found this Honeywell patent online:
http://www.freepatentsonline.com/7414546.html
So, I stand corrected - there ARE LED anticollision light systems available nowadays, even though they're relatively new.

My concerns are:
1) Safety.
2) The abilities of the original poster to understand electronic schematics, ability to create circuit boards, troubleshoot said boards, etc.
3) Attempting to design such a device remotely without actually building and testing it relates to 1), above.
4) Actual performance of the final system.
5) The time involved to design such a system.

This isn't going to be simple, nor particularly cheap. There should be built-in fault detection/monitoring, along with redundant LEDs. Heat-sinking the LEDs could get complex; they'll be dissipating a significant amount of power as heat, and they will need to be enclosed behind a clear lens/window of some material (Plexiglas, Lexan or the like).

For an aircraft that might rack up tens of thousands of flight-hours, the time and cost for developing such a system would be well worth the trouble.

However, (to be continued...)
[eta] (...continuing)
Xenon flash tubes have been in use for quite a while (longer than I've been around), they're really fairly simple to get flashing, don't have complex cooling requirements, and are plenty reliable for small (and large) aircraft use.

there are lexon as small as LDS devices and bright as the sun but they are not LEDS the current dissipation is into to 10-20 watts per device. I have one

 

I've already talked to a couple of the guys, and they're open to the idea. But we're talking about through a window, not opening one and sticking the light out.

One guy did offer to attach the light to his R/C helicopter, but I doubt he'd be able to get it up to 5,000 feet.

Having the devices on the ground, pointing up, to fly over would give a pretty good idea of visibility. If visibility is less than crisp while they are on the ground pointing up, and you know where to look, it will be a bigger issue when they are against the sky for background.

A large white area, relative to the size of the strobes, such as the roof of storage units, would be a good simulation on a sunny day for 'worse case' checking. Again, this is suggested to take place in an area where the test would not be mistaken for a distress signal, or any other kind of beacon. The narrow dispersion angle should keep issues away on the horizontal, and an aircraft would need to be looking to see it. I'm guessing the pilots know your local do's and dont's best, and might even have a better solution.

Trying them from the window, even with LEDs could cause flash blindness from the partial mirror of the plexiglass if they weren't held flush. In addition, the unit would need to be aimed at the observer for a longer time due to "aiming error", thus maneuvering the plane at attitudes/positions not fully condusive to continual flight.