Author Topic: How to design test methods that are repeatable & reduce experimental uncertainty  (Read 294 times)

agent_smith

  • Sr. Member
  • *****
  • Posts: 1382
I thought I'd start a new topic thread.
The title speaks for itself.

This is the IGKT and it is a forum for knot geeks.
So I think the topic title and intent fits within the scope of the IGKT.

The ten commandments for every aspiring youtuber knot tester

1. Don't indicate the test article material specifications (ie don't provide any details about the type of rope/cordage used in the test - just let people guess. All ropes are the same.)

2. Don't use a load cell to measure the magnitude of the force - just refer to it as a 'force'. A force is a force - who cares how much force? Measurement is for geeks.

3. If you use a drop mass to generate a dynamic loading event - don't specify the amount of mass used (its a guessing game - just let people guess the mass). Also, who cares about the magnitude of the fall distance? A fall is a fall - gravity sucks right?

4. If the focus and subject of your test is a particular concept - don't bother defining it. Just assume that your audience understands...

5. Mix two different knots in each test run (ie use different knot geometries to introduce more variance). Variance is good - the more the better.

6. Your mindset must only be 'pull-it-till-it-breaks' - nothing else is important. Strength is everything - and this must be your default mindset. Knot efficiency is only defined according to the metric of force. other factors such as the amount of rope material required to form the knot, whether the knot is resistant to jamming, whether the knot can also be tied without access to either end, the footprint (size/volume) of the knot, and its stability + security under a variety of loading profiles are all irrelevant and in the realm of science geeks and aliens.

7. If you perform a slow pull type test (eg with a ram on a horizontal test bed) - ignore the fixing at the stationary anchor - only observe what happens to the knot at the moving anchor point. If there is any pattern to which side the knots are breaking, don't report it - keep it confidential.

8. Never reveal exactly how you tied your knot specimens (eg don't take high quality photos/video of your knot specimens) - just assume that everyone knows the type of knot and how its tied (other people don't need to know exactly how you tied your knots - it should be obvious, so why waste time and energy trying to document it?).

9. Just do one test - statistically valid samples are for science geeks and aliens - one test is all you need to prove a point.

10. Don't bother using an experimental 'control' - the whole concept of a control is for science geeks. Real science can be done just as effectively without using a control.

...

If you successfully achieve all ten points, go ahead and upload your video to youtube and publish to the world.
Be proud of your achievement and contribution to knot science - sit back, wait for all the 'likes' and take delight that by default, the lay-public will perceive you to be an expert.

SS369

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 1970
I thought I'd start a new topic thread.
The title speaks for itself.

This is the IGKT and it is a forum for knot geeks.
So I think the topic title and intent fits within the scope of the IGKT.

The ten commandments for every aspiring youtuber knot tester

1. Don't indicate the test article material specifications (ie don't provide any details about the type of rope/cordage used in the test - just let people guess. All ropes are the same.)

2. Don't use a load cell to measure the magnitude of the force - just refer to it as a 'force'. A force is a force - who cares how much force? Measurement is for geeks.

3. If you use a drop mass to generate a dynamic loading event - don't specify the amount of mass used (its a guessing game - just let people guess the mass). Also, who cares about the magnitude of the fall distance? A fall is a fall - gravity sucks right?

4. If the focus and subject of your test is a particular concept - don't bother defining it. Just assume that your audience understands...

5. Mix two different knots in each test run (ie use different knot geometries to introduce more variance). Variance is good - the more the better.

6. Your mindset must only be 'pull-it-till-it-breaks' - nothing else is important. Strength is everything - and this must be your default mindset. Knot efficiency is only defined according to the metric of force. other factors such as the amount of rope material required to form the knot, whether the knot is resistant to jamming, whether the knot can also be tied without access to either end, the footprint (size/volume) of the knot, and its stability + security under a variety of loading profiles are all irrelevant and in the realm of science geeks and aliens.

7. If you perform a slow pull type test (eg with a ram on a horizontal test bed) - ignore the fixing at the stationary anchor - only observe what happens to the knot at the moving anchor point. If there is any pattern to which side the knots are breaking, don't report it - keep it confidential.

8. Never reveal exactly how you tied your knot specimens (eg don't take high quality photos/video of your knot specimens) - just assume that everyone knows the type of knot and how its tied (other people don't need to know exactly how you tied your knots - it should be obvious, so why waste time and energy trying to document it?).

9. Just do one test - statistically valid samples are for science geeks and aliens - one test is all you need to prove a point.

10. Don't bother using an experimental 'control' - the whole concept of a control is for science geeks. Real science can be done just as effectively without using a control.

...

If you successfully achieve all ten points, go ahead and upload your video to youtube and publish to the world.
Be proud of your achievement and contribution to knot science - sit back, wait for all the 'likes' and take delight that by default, the lay-public will perceive you to be an expert.

Just love tongue-in-cheek. 
Replace Don?t with Do and it hits the Mark.

I think it is time for a good list for the home enthusiast to follow so others can replicate and confirm the results, or (k)not.
I don?t see the need for absolute breaking tests here, most of us will never use a rope till it breaks. And there are the manufacturers specs to use as a safe use guideline.
The testing is not necessarily confined to specified brands, but, the materials should be specifically stated so as to be able to repeated by another.

I don?t see the need for a one knot versus another type unless there is a specified use of that particular combo.

Three tested samples of a determined configuration seems adequate. And clear photographs of specimens, before and after.

Perhaps someone has an idea of a simple, inexpensive, (readily available materials) rig design, ei. 2x4 lever with specified measurements that can be duplicated by anyone inclined to participate.
Or drop test with a known, available weight source (sand or water in a container?)
Anyone?

SS



Dan_Lehman

  • Sr. Member
  • *****
  • Posts: 4056
A key item is specificity of the knot geometry
--WHAT is being tested.
References to books are pretty useless,
as books are not specific (often, e.g., one
cannot even get which end is loaded!).

Maybe a specified setting load (say, around 25#
for 1/4" to 1/2" cordage?) and a couple photo'd
faces/sides of the knot.

As for having some number of test cases to satisfy
statisticians, I see this as a serious impediment to
making advances.  I will be happy with getting the
knot state and where-it-breaks shown clearly over
having NN cases in some single rope and trying to
make that information matter.


--dl*
====

KC

  • Sr. Member
  • *****
  • Posts: 403
Put Do in green and NOT in red?
Maybe with parenthesis,
End wit smiley in parens.?
« Last Edit: September 18, 2021, 01:53:21 PM by KC »
"Nature, to be commanded, must be obeyed" -Sir Francis Bacon
We now return you to the safety of normal thinking peoples.
~ Please excuse the interruption; thanx -the mgmt.~

agent_smith

  • Sr. Member
  • *****
  • Posts: 1382
With regard to measurement of force...

If you are intending to regularly test knots and upload your results to youtube - you should definitely consider purchasing a load cell.
However, load cells are not exactly cheap, and so you need to carefully think about what it is that you are trying to achieve, and what your overall goals are.

If you wish to remain in the realm of a hobbyist / backyard tester, you should at least make an effort to be consistent.
A cost effective solution is to use gym weights (refer to attached image).

Kettle bell weights are convenient - because they already have a cast attachment point.
They are commonly available  up to 40kg size... and it is possible to purchase even heavier (but price dramatically increases).

If using gym weights, it implies some sort of vertical test rig (gravity acts downwards).
You would need to be able to hoist your weight - without causing personal injury/strain.
If using disc weight plates, you would need to bolt them together with a long length eye bolt (bolt diameter between M20 - M30).

...

All of this of course comes down to whether you want your test results to be repeatable by others.
If you couldn't care less, then don't bother making any effort to accurately measure force.
But, keep in mind that your credibility may be impacted if you test in way that makes it impossible for others  to repeat your results.