Building industrial automated golf ball tester - Ideas?

[morphious]

HI Guys, Love the board!

I am in the process of building a new golf ball for a game I am creating and I need to test the golf balls I am receiving from China for consistency and durability. I am a start up company and do not want to spend $20,000+ to buy and industrial factory built automated system..... So that is why I am here!

I am looking to build a compressed air system, with all metal parts ( for durability ) and some kind of ball return system and a counter. So the idea would be to build an all metal cannon and shot the ball into a metal box. ( attached at the end of gun ). The ball should hit the plate at 300 (feet per second) = 204.545455 mph. Then the ball should be returned to the cannon ( maybe have it roll back in ? ) and it then fires again. I hope to get 1000 shots per golf ball. So this would be running for days on end repetitively.

Would need a way to count the shots, speed, etc. ( I have seen many on here measure speed etc, so advices on this would be Excellent, But must be able to be industrial strength )

BTW here is the link to the $20k + one ( price goes up to 300K ) http://www.automateddesign.com/node/6

So ideas on parts, compressors for rapid fire, box ideas, return methods, etc would be HIGHLY appreciated.

-M

[POLAND_SPUD]

hmm that seems doable
3/4" QEV piloted by a solenoid 3 way, an air cylinder + arduino and some sensors to measure velocity

[velocity3x]

The ball should hit the plate at 300 (feet per second) = 204.545455 mph.

Easily doable project.....

[jimmy101]

Photodetectors would certainly work but a MIC plugged into a laptop is probably all that is needed for the chronometer.

The question I would have is exactly what you are trying to test?

Consistence? Mass, minimum and maximum diameter (most "spheres" are actually ellipsoids). Those are easy enough to measure. Coiefficient of "rebound" (wrong term, the correct one escapes me at the moment). Just drop the balls from a fixed height and measure the height of rebound. Photo diodes would work, or again, a MIC + laptop could measure the height of rebound based on the time from "plink" to "plink" as the ball repeatedly bounces off the plate. A little external ballistics calc would give the coefficient of "rebound". Velocities are low enough that air drag can probably be ignored.

In an air gun setup I would think that most of what is going to change is the balls getting smashed when the hit the stop plate. The design Poland posted won't put much stress on the balls at launch. It'll be the smashing into a steel plate at ~200 MPH that'll "wear out" the balls. If you want to mimic the strike of a club you would probably need to add a piston to Polands design. Then, use something much softer to stop the ball down range, perhaps several heavy canvas curtains?
Edit: Or reorient the gun so it fires straight down and shoot into a water tank? Recycling the ball will be a bit tricker.

[POLAND_SPUD]

but a MIC plugged into a laptop is probably all that is needed for the chronometerbut a MIC plugged into a laptop is probably all that is needed for the chronometer

I suggested areduino because it's cheap easily available and very well documented.
What is more, he could control the gun and measure stuff at the same time (an extra optical sensor that checks if the ball is ready to be lifted by the air cylinder etc.).
Finally, it wouldn't be very difficult to send the results over the USB cable, which acts as a virtual COM port. There is even a simple example in processing that shows you how to read data from serial and export it as csv or xml file with all headers and sh## you want

ohh and you can even add a few lines of code that will send results to a website or just display info on a website whenever the machine stops working the way it should (like for example if the ball doesn't return to the lift for any reason)

[Technician1002]

What he is trying to measure is a very valid question. If the ball is tested for loss of firmness indicating a breakdown of the inside of the ball, a steel plate with a magnet and pick up coil can be used as the target. If the ball has an elastic collision, the transfer of momentum to the steel plate will be high. An inelastic collision will be lower energy. Using launch velocity and mass to calculate the launch energy, the energy gain of the plate can be measured so degrading over 1000 shots and impacts can be quantified. In golf, this would be seen as a reduction in range with ball age. An alternative to spring mounting the strike plate, the plate can be mounted very solid and a photo gate can measure the ball velocity into the plate and again away from the plate and energy loss can be measured as delta velocity. A non bouncy ball will have much lower exit velocity. A really good ball will exit at nearly the same speed as the approach. The change over a series of shots will show the rate of decay of the ball.

[morphious]

Hello and thank you all for your input so far! The goal of this is to test the COR and to test the durability of the ball itself. We want to see when the ball looses compression to the point where it is unusable, or the outside shell is destroyed to not be usable either.

We intend to test multiple high quality well know balls, then compare the balls we are getting from China to the results of what we received from the high end balls. Since we would order a large order, we would test a few per batch to make sure the quality, degradation, etc remains the same or better then our original tests.

Because of new materials and design changes in most balls over the past 2 years, a lot of manufacturers are saying they can get 1000 normal strikes before the ball is unusable. They use the machine listed above to test. And this is between 140-180fps. This is using what looks like a piston to hit the ball into a soft surface. I would do this too, if I could find that piston But the other cannon one is set to 300 fps and shot at a club head. At 300 fps, into a metal plate is good, This is the MAX an golfer can hit the ball. This would be an overkill test as the, and if we use the same psi for all balls, we should be able to get consistent results to be able to test the good ball data against the balls we retrieve.

OK, so now that we all understand the data I am trying to gather, lets move on

POLAND_SPUD: Cool, your drawing makes seance. Going to have to have a Strong back plate, to hit. But good job with this..

I am amazed at the amount of speeds of the golf balls that you are all shooting on here, Insane! My application , obviously calls for a Lot Less then what most people shoot out of their cannons!

Any more help would be appreciated!

[morphious]

Technician1002 - This is all very good, but there is an easier way to test this. One is a visual inspection of the outer shell. ( not very scientific but most golfers will not play with a ball that is really messed up looking... ) the second easy way if to measure the compression of the ball. The compression of the ball will drop as the materials start to decay due to the abuse.

No golf balls would go though this in real life situation, as most people loose them before 10+ rounds of golf is used, plus the only club that would come close is a driver, and 95% of golfers cannot hit the ball with that amount of speed. So most of the balls use is Much lower ten a 200 mph strike with a driver.

So this test would be a quantified test.... ( way over kill )

[velocity3x]

Use a PLC for process control, data collection and communication......like real-world, industrial equipment. $119

DL05 CPU, 8 DC in / 6 Relay out, 110/220 VAC power supply. 6k words total (2048 words ladder - flash, 4096 words V-memory), RLL/RLLPLUS programming, built-in RS-232C programming port and additional RS-232C communications port. Inputs: 8 DC inputs, 12-24 VDC current sinking/sourcing, 2 isolated commons. First 3 inputs are configurable in one of several high-speed I/O features such as 5kHz counter input, pulse catch input, or interrupt input. Outputs: 6 relay outputs, 6-27 VDC, 6-240 VAC, 2A/point max., 2 isolated commons. One option slot available for I/O or communication module.

[morphious]

That is a really nice controller, if I only knew how to user it I guess I will have to read more on the chat board.

[velocity3x]

That is a really nice controller, if I only knew how to user it I guess I will have to read more on the chat board.

They're easy to use.
Basically.....You connect the PLC to a PC via cable.
Run the programming software in the PC.
Fill in the blanks of the programming software with the functions you select from menus.
Press 'SAVE" to send and store your finished program in the PLC.

[D_Hall]

My take on the topic....

Vertical gun with a cage mounted to the muzzle. The cage's bottom is roughly funnel shaped.

Gun fires. Ball goes up into cage and bounces around. Once energy is expended it falls to the bottom where the funnel shape drops it back into the barrel. It falls to the bottom where it awaits being fired again.

Play around with it a bit to see how long this process takes. If it takes X seconds at the max....

Then simply attach a timer. One shot every X+5 seconds. Nice part about that is a clock will tell you how many shots you've fired. No fancy shot counter required.

For extra credit: Make the last 12" of your barrel (or the first 12" of your funnel) out of plexiglass and you can use a commercial printing chronograph to get shot data for every shot.

[shardbearer]

The vertical gun is just duh. You would need a bit more power, but that shouldn't be a problem. Much easier than having to deal with breeches and air cylinders.

And I would use an Arduino for both D_Hall's method of putting the shots on a timer and for recording fps for each shot. You could even look at the data afterwards, and see, for example, that the fps dropped suddenly at x shots because the outer shell broke and fell off. And instead of having a clear barrel (impossible because you can't buy clear thinwall pvc) or a clear funnel, drill four holes in the funnel (where it doesn't need to be airtight anyway) for the leds and phototransistors. Don't forget to put big air escape holes in the funnel too.

And you need a solenoid triggered 3 way valve. Might be hard to find. Hell, have a standard solenoid hit the button on a standard 3 way valve.

[D_Hall]

And you need a solenoid triggered 3 way valve. Might be hard to find. Hell, have a standard solenoid hit the button on a standard 3 way valve.

Nothing hard to find about a solenoid triggered 3 way. I buy 'em by the dozen at the office.

[morphious]

D_Hall - that is Great! That is Exactly the idea I was planning on doing. I am glad some one else came up with this idea too!

Build the cannon so it points up at 45 degree angle, have a box at end that is horizontal, fire the ball into a steel plate in the box, and then have a v mounted at the bottom of the box, so it guides the ball back into the cannon barrel to reload. When it hits the bottom on the barrel it trips a sensor that then counts down to 5 ( let it settle ) then fires again.

A simple PLC with HMI and lcd screen can be set up to start and stop the process and count up the amount of shots.

I think this is the simplest and most efficient way to do it and I do like the idea of vertical system, this seems to be the best way to return the ball, with the least amount of effort.