Hip Pocket Builders' Forum

Indoor Free Flight Forum => Indoor Building Tips and Problem Solving => Topic started by: dslusarc on February 18, 2017, 09:24:38 PM



Title: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 18, 2017, 09:24:38 PM
I wanted to test several batches of rubber I have and started to do my tests using the Fred Pearce method and have run into an issue with his formulas with modern rubber. His method involves doing an initial pull and then a second pull to take the data. The second pull is based on the length of the first pull. The problem I have is the formula for the second pull gets me a lower force value than the first pull. It is suppose to be slightly higher. The reason this happens is looking at his formulas, he assumes rubber can not exceeded a 9.555 : 1 stretch ratio.

Pearce Method Formula's

F1 (pounds) = 45 x weight in grams / initial loop length

So F1= 45* .480 grams/5.8" = 3.72 pounds

I pull the rubber to 3.72 pounds and that measurement (L1) was 56.875 inches. I let the rubber sit pulled at that distance for 5 minutes, then step #2, you pull to a slightly higher force based on initial pull.

F2= 430 * weight in grams / L1

F2= 430* .480/56.875 = 3.63 pounds which is less, hence the problem I have.

F2 is suppose to be greater than F1, so rearranging the formulas I got

L1 < (430/45) L

or L1< 9.555L

or L1/L< 9.555

Well at F1 my L1= 56.875 so my stretch ratio is 56.875/5.8= 9.806 which is not less than 9.555

The factor of 430 Fred derived was for older Pirelli/FAI black rubber. For tan rubber with higher stretch ratios over 9.555 this 430 value needs to be increased, but I do not know how he determined the value initially so not sure what to change the value to.

Here are some examples from a letter he sent many years ago on indoor rubber testing.
Chilton Pirelli 7/13/80
length= 16",  weight" 1.5309 grams
F1= 4.305 pounds, L1=111"
F2= 5.93 lbs   L2=120.5"
stretch ratio= 7.53

78 Pirelli 7/8/80
length= 16.25",  weight= 1.063 grams
F1= 3.183 pounds, L1=124"
F2= 3.68 lbs   L2=132.5"
stretch ratio= 8.15

FAI Black  7/6/80
length= 10.25",  weight= .765 grams
F1= 3.36 pounds, L1=72.75"
F2= 4.524 lbs   L2=76.5"
stretch ratio= 7.46

I am trying to correlate these three data points into a way to adjust the "430" factor in the F2 formula. Any thoughts or any input on how to determine the pull force to take data?

Don



 




Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 20, 2017, 03:08:26 AM
I think I have figured out how to make my adjustment. Fred spoke of pulling the rubber to a constant stress. Using that idea I see he was going to a stress of around 6500 psi. I have adjusted to around 7000 psi so in my F2 formula I use a factor of 480 instead of 430 and I was able to successfully test all 6 batches of rubber I had today. My knots kept pulling through on the 5 minutes rest, so had to go to three knots then it stopped. I was using two knots like I usually fly with but after 2 minutes at F1 force they would slip. So adding one more knot did it.   

for stress I used this formula based on constant volume:

stress=  F * (dL +L) / (A * L)

L= initial length so
dL +L = stretched length
A= original cross section area before stretch so starting width * thickness and since two strips , area is doubled, multiply by 2

I then got:
stress= (Force * stretch length )/(2 * starting thickness * starting width * starting loop length)



My test results for each batch are as follows.

Batch   Energy(ft-lbs/lb)   Stretch ratio

8/93     3992     10.52
10/97    3941     9.66
5/99     4109     10.26
9/14     3851     9.07
3/16     4000     9.14
6/16     4042     9.24

where stretch ratio : stretched length at F2 force reading /original length

I forgot to test my 3/02 rubber so will do that next.  I have plotted the raw data on the first graph. To compare the curves to each other I normalized the actual lengths to stretch ratios and the force I divided by the rubber weight. Not sure if that is a useful way or not. I wanted to compare the curves to each other. 

Don



Title: Re: Question on the Fred Pearce method for testing rubber
Post by: duration on February 20, 2017, 06:57:01 AM
Don,

I found some old numbers that Fred Pierce sent me some years back. The only overlap with the batches you tested was October 1997; his range was 3900-4060.

Here are the other test results:

Sept. 96 3770-3830
Nov 96  3540
June 97 3440-3750
July 97 3910-4190
Oct 97 3900-4060
Dec 97 3840-4180
Feb 98 3870-4007
May 98 3920-4030
July 98 3760-3780

There was no additional info on temperature or stretch. One of the times I visited Fred he showed me the testing "rack". It was outside, but protected by a roof overhang. It looked like something you would see at a historic farm recreation. I believe that Fred adjusted/corrected testing results for temperature.

Louis


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 20, 2017, 07:27:04 AM
His getting a similar resuit on the same batch is encouraging to me. Thanks!


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: duration on February 20, 2017, 08:45:03 AM
Don,

Glad to see your numbers for 6-16; I picked up 5 pounds last summer and just received an additional 5 pounds. Looks close enough to May 99 for government work.

Louis


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 20, 2017, 09:40:03 AM
When I heard rumors at the indoor Nats last year I bought some 6/16 and some 3/16 batches over the phone in the parking lot :)


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: jakepF1D on February 20, 2017, 11:48:31 AM
I have no experience with 6/16 or 3/16, but I do have some Tan SS from a few years ago that I've tested.  The problem with the Tan SS in my experience is not the total energy it can store, but where it stores that energy.  I'm able to wind the SS rubber to much higher torque values for a given length and weight of motor, but I can't use all that torque at the very end of the wind.  This is why 5/99 is so appealing.  It doesn't reach the same peak torque, but it maintains torque in the middle of the curve better than any other batch.  This results in a manageable climb, and a nice long cruise.  Perhaps this has changed with the newest batches, but my understanding is this torque curve is intentional because it benefits the outdoor flyers that use most of the rubber.  Time will tell I guess.


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 20, 2017, 12:16:10 PM
The pull curves seem to confirm  that observation. I have been helping some kids with So models and using the 9/14 batch. I notice the end of the flight the models drops real fast on those last winds. I am encouraged about the 8/93 tests I did as I have a fair mount of that rubber left. Been stored in mason jars in a freezer for 24 years. It seems closest to the 5/99 curve.


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: Tapio Linkosalo on February 20, 2017, 12:37:39 PM
I do not think the torque curve of SS is such for intend, but it is the characteristics of the rubber you get with the current cocktail of add-ins. (and they cannot make Tan II any more, as some add-ins have become obsolete).

I wonder how the 2016 batches of SS would look if the motors were a tad longer, so that the "cruise torque" would overlap that of the 99/5?

BTW, the pulling test results agree with those I have gotten with my tests of F1B motors using the Pearce method.

 


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 20, 2017, 12:52:42 PM

I wonder how the 2016 batches of SS would look if the motors were a tad longer, so that the "cruise torque" would overlap that of the 99/5?


I was thinking the same thing.  Slightly longer loop the same weight and plotting torque curves to see the overlap. When I first started indoor I used Pirelli and FAI black. When tan came out motors were a lot shorter than with Pirelli for same turns. I went from using 17-18" loops on F1D to 15-16" loops at the time. Still have some of that rubber as well. I do have the actual raw pull test data from Fred for the three batches I posted in the first post (78 pirelli, chilton pirelli and FAI black) it will be interesting to see those on the chart to compare as well.

Don


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: duration on February 20, 2017, 01:56:01 PM
Don,

Here are some more numbers from Fred Pierce:

1978 Pirelli (Cannizzo) 3910, 7.75 stretch "brittle"

1978 Pirelli (Bob White) 3680 8.1 stretch "not brittle"

1976 Feb. FAI 3200-3600 7.1--7.3 "variable"

All I have the photocopy of a page, probably from an old Sympo.   Rubber samples run from 1976 to 1983. Lowest number in the tests was 3020, best was 3740.

There is also a note:  "Energy storage varies almost linearly with temperature. For example FAI rubber measuring 3400 foot pounds at 75 degrees F would measure 3070 at 50 degrees F and 3730 at 100 degrees F."

Louis


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: ram on February 20, 2017, 03:51:48 PM

I wonder how the 2016 batches of SS would look if the motors were a tad longer, so that the "cruise torque" would overlap that of the 99/5?


I was thinking the same thing.  Slightly longer loop the same weight and plotting torque curves to see the overlap. When I first started indoor I used Pirelli and FAI black. When tan came out motors were a lot shorter than with Pirelli for same turns. I went from using 17-18" loops on F1D to 15-16" loops at the time. Still have some of that rubber as well. I do have the actual raw pull test data from Fred for the three batches I posted in the first post (78 pirelli, chilton pirelli and FAI black) it will be interesting to see those on the chart to compare as well.

Don

I was also thinking along those lines, but thought I was visualizing it wrong.  Makes me feel a little better!!

On a side note, has there been comparable testing on 7/99? 

Rey


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: jakepF1D on February 20, 2017, 06:57:25 PM
Somewhere I have notes on some rudimentary rubber testing I did in mid 2013, but I remember the basic conclusion I reached was that I wasn't able to produce a torque curve with the Tan SS that matched the 5/99 while also matching the turns.  I could either get similar turns with less usable torque (my F1Ds are built and trimmed to handle 0.5oz/in of launch torque, anything beyond this is of questionable usefulness), or similar usable torque with fewer turns.

This may be different with newer batches, but I suspect the results may be similar considering the formulation likely hasn't changed much in the last 4 years.  I may also try this testing again now that I have better winding techniques and more knowledge about maximizing the output from my motors.


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 20, 2017, 08:16:03 PM
I wonder if the VP prop can be tuned to the TSS rubber. I can see a model being optimized for 5/99 then swapping rubber it may not fly right as the prop is tuned for 5/99 curve. I know when I was flying 65cm and went to 5/99 rubber from 10/97 my models would climb through the roof at Akron. On 10/97 I would get say 15-16 minutes of climb but 5/99 was 17-18 minutes of climb and I had to make a flaring prop for Akron to keep off the ceiling with 5/99. I wonder if a person took one of the TSS batches and tuned the VP prop,  spring, preload etc for TSS specifically if it would work out?

Don


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 21, 2017, 10:24:08 AM
Looking at the stretch ratio I obtained for the 9/14 batch I have been using with the SO kids, the tested valued by pulling is 9.07 stretch ratio. I have been using a K factor of 42.1 in my max turns calculation for that batch based on winding to break point. Using the 9.07 value and converting units that equates to

9.97 x 2 / 13/.035275 = 42.85 which is pretty close. Or going in reverse, 42.1x12x.035275/2= 8.91 stretch ratio. So it correlates pretty well.

Don


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: Tapio Linkosalo on February 21, 2017, 10:35:51 AM
I wonder if the VP prop can be tuned to the TSS rubber. I can see a model being optimized for 5/99 then swapping rubber it may not fly right as the prop is tuned for 5/99 curve. I know when I was flying 65cm and went to 5/99 rubber from 10/97 my models would climb through the roof at Akron. On 10/97 I would get say 15-16 minutes of climb but 5/99 was 17-18 minutes of climb and I had to make a flaring prop for Akron to keep off the ceiling with 5/99. I wonder if a person took one of the TSS batches and tuned the VP prop,  spring, preload etc for TSS specifically if it would work out?

I think you still could not use all that high torque energy of TSS, as VP-props tend to loose in their efficiency at high pitch settings.

But on the other hand, TSS has a flatter cruise curve. Maybe it should turn out to be superior for fixed pitch prop classes?


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: jakepF1D on February 21, 2017, 10:37:25 AM
I wonder if the VP prop can be tuned to the TSS rubber. I can see a model being optimized for 5/99 then swapping rubber it may not fly right as the prop is tuned for 5/99 curve. I know when I was flying 65cm and went to 5/99 rubber from 10/97 my models would climb through the roof at Akron. On 10/97 I would get say 15-16 minutes of climb but 5/99 was 17-18 minutes of climb and I had to make a flaring prop for Akron to keep off the ceiling with 5/99. I wonder if a person took one of the TSS batches and tuned the VP prop,  spring, preload etc for TSS specifically if it would work out?

Don

It's certainly possible to tune the prop to make it work, but ultimately the area under the curve will determine how much total energy is available.  More energy will result in a longer flight time assuming the model and prop are properly trimmed.  If you can build a model that can hold 0.6-0.7oz/in of torque without breaking, and utilize that torque burst during launch, you may be able come very close to 5/99.


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 21, 2017, 11:51:36 AM
I am just thinking out loud here, the 6/16 and 3/16 batches both test equal or slightly higher than my 10/97. Now I use 10/97 all the time. I always felt it was great rubber for under 100 feet. Less climb but great cruise. I would think that in theory at least I should get equal time with those two TSS batches as they have equal energy. I may need to make a prop change to do it but I would think it would be possible. As far as .7 in-oz torque, we use to do it all the time on 1 gram 65cm F1D . I use to fly 14" loops weighing 1.4 grams. 2050 turns minus 40 was around .7 in-oz if I recall correctly. So it is not unprecedented.
Don


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 21, 2017, 09:11:06 PM
OK tested 4 more batches of old rubber, they are the first four on this revised list. 

Batch   Energy(ft-lbs/lb)   Stretch ratio

5/96     3829     9.74
6/96     3583     9.66
10/96   3975     9.45
5/97     3864     9.66

8/93     3992     10.52
10/97    3941     9.66
5/99     4109     10.26
9/14     3851     9.07
3/16     4000     9.14
6/16     4042     9.24


The second graph I zoomed in and removed a few of the other batches, it is interesting to see the 5/96 batch is closer to the 5/99 curve until you get past 8:1 stretch ratio. One more batch to test, the 3/02. 


Don


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 22, 2017, 12:10:47 AM
OK my tests on the 3/02 batch are completed. I am surprised at the results but in retrospect it makes sense. I always thought my 3/02 was pretty good rubber and now I know why, it has the highest total energy of the batches I have.   

Batch   Energy(ft-lbs/lb)   Stretch ratio
3/02     4142    9.75

5/96     3829     9.74
6/96     3583     9.66
10/96   3975     9.45
5/97     3864     9.66
8/93     3992     10.52
10/97    3941     9.66
5/99     4109     10.26
9/14     3851     9.07
3/16     4000     9.14
6/16     4042     9.24

Here is a chart showing the 3/02 compared to a few select other batches. Interesting it is like 10/97 near higher stretch, but has higher force the rest of the way down.

Don


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 22, 2017, 12:18:32 AM
Here is the 3/02 loop retested and a comparison made from my normal testing pace (about 4-5 minutes to collect all data points) to waiting to let it stabilize on my scale at each length (about 20 minutes to take all readings). I found that at the higher stretches it took longer to stabilize, about 2 minutes then got shorter as the force/stretch was less and less. The graph also has the actual force differential seen in my two collection methods. The regular pace showed an energy of 4142 ft lbs, the slower pace showed a energy of 3905 so 94.28% of the initial tested value. I am in process of doing the same test to the 6/16 TSS batch.

Don

 


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: Olbill on February 22, 2017, 10:15:28 AM
The 3/02 you sold me several years ago has won contests and set at least one record for me. The only problem is that it's all gone (except for a few scraps)!


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on February 22, 2017, 09:29:02 PM
Here is todays test on 6/16 TSS broken in then pull tested at my normal rate then waited one hour and pulled again to the same distance and then held until the readings stabilized at each point. Typically 1-2 minutes at each location , as before, the higher stretch levels had to be held longer to stabilize. The results on this new test was interesting:

Fast release (3 minutes): 4106 ft-lbs
Slow release (20 minutes): 3775 ft lbs

3775/4106= 91.94%

Compared to the 3/02 which went from 4142 to 3905, or 94.28%

So the TSS gives up more energy on slower releases.

Don
 


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: Olbill on March 02, 2017, 01:54:39 PM
The 3/02 you sold me several years ago has won contests and set at least one record for me. The only problem is that it's all gone (except for a few scraps)!
I found a 12' length of this rubber in my scrap bag so made (3) LPP motors around 2.7g x 23". The three off cuts made a 6 strand motor of the same weight and 2 other scraps made a 4 strand motor around 20" long. I'm a happy camper!


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: dslusarc on March 02, 2017, 02:08:49 PM
Excellent! How do you like multistrand indoor motors? One of the reason I fly models like ministick is from rubber cutoffs. Never really thought of multistrand that much.


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: msc on March 11, 2017, 09:28:00 AM
Back to the original question. Looking at the test methodology Fred Pearce published in 1979 I have a couple thoughts on selecting the F1 and F2 forces.

I expect the 45 used for F1 and the 430 used in the F2 equations are empirically derived. Pearce states the first pull is to stretch to near the maximum stretch and the second pull is to bring it near the breaking point. With that in mind I would do a few test and derive new factors for both the F1 and F2 equations.

I suggest you test a few samples the way you have been then do one more pull much like the second pull but keep going until you find the breaking point. With this data I would determine a factor that typically gives an F1 force about half way up the final steep slope and a factor that typically gives an F2 a bit short of the breaking point. If you make the third pull to destruction a regular part of the test procedure you add breaking point to the accumulated knowledge and you get feedback on how close your second pull is coming to the breaking point. Update the factors as necessary so the F1 and F2 guide numbers put you where you want to be on the curve.

Mike


Title: Re: Question on the Fred Pearce method for testing rubber
Post by: duration on March 31, 2017, 01:59:36 PM
I just stumbled across yet another way to test rubber. "A Quick Rubber Test" by Hank Cole ran in the May 1997 issue of the "Bat Sheet".  It had been filed away years ago; I discovered it going through some old articles on rubber.  The title is right--it is a very quick test; here are the steps; (I've included numbers I got from a test of June 2016 Super Sport.)

* Measure out exactly 12 inches of rubber, weigh on accurate scale and note the weight (wt = .73)

* Mark and accurate 1 inch section in the middle of the 12 inch strip and pull to limit and measure the max stretch (S = 9.5)

* Subtract 1 from the stretch number and multiply by 0.58. (9.5-1 = 8.5 x 0.58 = 5.93) This is the pull-to number.

* Tie loops in both ends of the 12 inch strip. Hook one end to a firmly-anchored peg or hook; fasten the other end to a pull scale. Position a ruler under the rubber strip and start pulling until the distance between the two marks on the rubber are at the pull-to distance. Note the force needed to reach the pull-to length in grams; this is the weight W.  (W = 353.8g)

* Multiply the weight W by the stretch length S-1 and dived that by the rubber weight wt. This gives the energy E in foot-pounds per pound. ( 353.8 x 8.5 divided by 0.73 = 4119)

I did my tests at 66 degrees F and used Paul Roster's temperature correction factor described in his 2016 Sympo paper (1.3% per degree C).  That changed my 4119 figure to 4178 ftlb/lb at 20 degrees C. (Paul did not test June 2016, but a recent numbers I have seen elsewhere for June 2016 is 4106.)

While I had Paul's Sympo paper open I checked his number against some of the batches of Super Sport that we both had tested. Here are the results for the only two overlaps:

June 2013  Cole method by Louis Joyner (3964)  Paul Rossiter (4050)

April 2015  Cole method (4715) Rossiter (4720)
(All energy results corrected to 20 degrees C)

The two problems I had with the Cole method of testing were trying to estimate the exact middle of the marks on the rubber (the two marks made an inch apart). When the rubber is stretched the very narrow mark made with an ultra fine point Sharpie becomes a wide smudge. Also both marks move away from the anchor as the rubber is stretched out to the pull-to distance; I had to constantly move the ruler to keep the zero mark directly under the center of one of the marks on the rubber. All the tests were done with 3/32 strip, but Cole does not specify strip width.

I have no idea how Hank developed this test but I found quick and easy, and, surprisingly accurate. It is, as with any test, dependent on the accuracy of the test equipment (digital weigh scale and digital pull scale) and the test operator.

Louis

PS  I also tested November 2012 and came up with 4374 corrected.