Testing the strength of heavy duty caving belts

Posted on 5th September 2017 by Pete

A Method of Testing the Strength of Heavy Duty Caving Belts

The aim of this was to establish a method to test the strength of heavy duty caving belts that did not rely on having access to a load cell. I hoped to produce a simple system that needed very little equipment and that would deliver a test load to a belt that exceeded the minimum strength requirement for its use.

Why? Well I felt that I needed to have some kind of empirical justification to use an item of non-PPE equipment in the role of a height security device. I didn’t feel that “because we have always used them in this way” was a sufficient argument for their use. As far as I’m aware there has never been a failure of a caving belt that led to an accident, but that is not really a reason for never questioning their use in this role. These are my personal thoughts and it does not constitute ‘advice’ or the position of the BCA Training schemes.

A quick note on use of belts – The user should never be in a position where they can become suspended on a belt alone. Additionally, they must never be subjected to falls or dynamic loads. They are for restraining movement to keep someone away from a fall hazard or preventing a slip becoming a fall on easy angled ground. They are no substitute for a harness where suspension is possible.

What strength does a belt need to be?

Well, this one is a potential can of worms…. Let’s be clear, the manufacturers do not condone the use of their heavy duty belts for taking any load at all beyond hanging your battery or lunch box from it. There is a historical use in cave and mine exploration that involves using the belt for the purpose of slip prevention and security on steep ground when combined with a rope belay or cowstails. If you were intending to use it for this purpose, especially as a leader of others, you’d need to be 100% sure that the belt was strong enough for that role. The manufacturers do not state this type of use is approved or list any strength rating on the product or the literature accompanying it. You must conduct your own test and risk assessment if you are to use them in this way.

If you want an item that has a certified standard for this type of use, you could choose to use a climbing harness, caving harness or potentially an EN358 work positioning/restraint belt.

For anticipating loads that could be applied to a belt in use, I have used a mass that is comparable to the maximum user weight ratings on some of the common PPE equipment at the time of writing: 120kg (Mass)
The caver has a short dynamic rope lanyard of 50cm length, fixed from their belt to an anchor. As discussed above, the user should never be subjected to a fall or suspension, but I am using the forces that it is possible to generate in ‘foreseeable misuse’ as a starting point for considering how strong a belt needs to be.
If they climb above the anchor, until the lanyard is tight, then ignoring all stretch or slack in a system, a possible FF2 fall of 1 metre can occur.
This FF2 fall will likely result in injury and, as a rule, cavers avoid putting themselves in a position where this kind of drop can be taken. By not climbing above the attachment point of there lanyard, the resulting fall cannot exceed FF1, or 50cm in this case.
When using dynamic rope cowstails, the UIAA standard permits stretch up to 40% of original length. For a 50cm cowstail, this is 20cm, or 0.2m (Impact Distance).

For a Fall Factor 2 (1m drop on to 0.5m cowstails)

velocity = √ (distance x acceleration due to gravity x 2)

v = √ (1 x 9.81 x 2)
v = 4.43 m/s

Kinetic energy = 0.5(mass x velocity²)

Ke = 0.5 (120 x 4.43²)
Ke = 1177.5 Joules

Impact force = Kinetic energy / Impact distance

IF = 1177.5 / 0.2
IF = 5887.5 N

Impact Force = 5.89 kN

This is clearly a very serious amount of force and is only a hair under the threshold that the work at height industry uses as a maximum safe force the human body should be subjected to. An impact of around 6kN on the body will cause injury in a lot of cases and should certainly never be taken on a heavy duty caving belt. It is beyond anything we should ever do when wearing belts and is included only to demonstrate the risk of improper use. A FF1 drop is still something to be avoided, but is more realistic of a potential real world scenario.

For a Fall Factor 1 (0.5m drop on to 0.5m cowstails)

velocity = √ (distance x acceleration due to gravity x 2)

v = √ (0.5 x 9.81 x 2)
v = 3.13 m/s

Kinetic energy = 0.5(mass x velocity²)

Ke = 0.5 (120 x 3.13²)
Ke = 587.8 Joules

Impact force = Kinetic energy / Impact distance

IF = 587.8 / 0.2
IF = 2939 N

Impact Force = 2.94 kN

So a 0.5m drop on to a 0.5m dynamic lanyard may produce a force of around 3kN for a 120kg caver. This does not take into account any stretch or bounce. This figure seems pretty reasonable, but we should seek more evidence to reinforce this for our follow up testing.

When considering the use of caving belts, can we can compare it to something done in another industry? Well yes, work restraint systems often make use of padded restraint belts instead of harnesses. One of the critical requirements for this system is that a user may not be permitted to go into suspension on this system. That seems very close to how we should be using heavy duty caving belts. When consulting BS8437 – Code of practice for the selection use and maintenance of personal fall protections systems and equipment for use in the workplace, we can identify that restraint belts need to conform to EN 358. Accessing this standard is expensive and no doubt the items conforming to this standard will have a very high safety factor. What we can get from BS8437 is the recommended strength of anchor points for use in a work restraint system. This is 3 x the mass of the user. A correctly installed and utilised work restraint system is only required to have an anchor of 3 x users mass. For our 120kg caver, this would be 360kg, or 3.6kN in force.

For our 120kg fictitious caver, we can mathematically predict a theoretical force of just under 3kN for a FF1 drop. We can also see that and anchor of 360kg (3.6kN) would be required if using similar techniques in work restraint. The figures are not exactly a match, but are comparable. Taking the worse case figure is probably the safest option going forward, so our belts must be capable of taking a force greater than 3.6kN for a scenario that does not involve wildly inappropriate use.

Safety Factor?

Apply to this any safety factor you wish. The 3kN figure from the maths is indicative of the maximum possible force generated in a FF1 drop on 50cm cowstails, the real world figure will be far lower due to stretch and slippage of the belt on the body and the sagging of the rope the caver is connected to. The BS8437 figure is a 3 x safety factor over the user’s mass anyway. You could argue that belts tested to 3.6kN would be sufficient as an indicator of appropriate strength if you never operated with cavers heavier than 120kg.

Belt Strength

Accepting all this, we are left with the figure of 3.6kN as our chosen minimum requirement for the strength of the heavy duty caving belt for any user we might encounter regularly (3 x 120kg based on BS8437).

So as long as we can apply a test force of 3.6kN or more to the belt, we can be assured that the item can hold the greatest possible force we can apply to it in proper use. The only remaining factor of concern is that would applying this force in test render the belt unsafe to use again, in essence, are these tests destructive? Only 1 way find out…..

Testing

Using 1 very large Corsican Pine and a good sized Birch tree, we set up a pull testing rig with a simple 3:1 theoretical configuration. I used a Rock Exotica load cell to get live feedback on the testing here but if you copy the method, you would not need to use one.

For the estimation of test force we regarded each person capable of pulling 50kg (see Gethin Thomas’ work on tyroleans). Through a theoretical 3:1 MA system that would be 150kg per person. With 5 undertaking the pull reaching 750kg and 6 equalling 900kg or approximately 7.5kn and 9kN respectively.

Kit used (minus load cell): Petzl rescue pulley, Petzl Basic jammer, Petzl Partner pulley, Lyon wire sling for tree, assorted karabiners, 20m rope.

Due to the force expected to be placed on the rope, I did not anticipate that I would be able to untie the end knot (fig 8 loop). This was accurate and the knot had to be cut from the rope end. Bare this in mind with your own rope!

We also used a Petzl Rollclip to redirect the angle of pull to make it easier to stand on the tarmac of the road alongside the trees.

Initially we had 5 people pulling the first test on a Lyon roller-buckle belt (brand new).
This produced a force of 5.9kN with no damage or slippage. This is lower than expected but there was a lot of tightening in the knot and stretch in the rope coupled with a general timidness of the pulling team.

The remaining tests used 6 people to pull. This one was conducted on my 10 year old Caving Supplies square buckle belt (already retired). This belt has nicks, fluff and rust and comfortably took a force of 7.74kN showing no damage or slippage. Next came my current AV belt, with it’s central maillon removed and directly attached to the pull line. This belt held 7.7kN without failure or slippage. Finally, the pulling team seemed at their most confident that nothing was going to break and send shards of metal and wood at them so they really gave the last belt some pain. This Warmbac square buckle belt was subjected to 8.64kN with no damage or slippage noted at the time.

It is not surprising that the force exerted by the pulling team was less than the theoretical 3:1 system implied. In practice with the loss of friction due to bearings and turns in the rope a 2.5:1 is a more real world figure and so our 5 x 50kg pulling average adults could be expected to make 625kg/6.25kN using this system.

On this test we pulled the belts to a far higher force than would be needed in a periodical strength test to simply demonstrate that this lower level of testing would not damage the belts. Using 4 people to pull on a 3:1 MA (2.5:1 actual) system in a reasonable way with un-gloved hands, would produce a force exceeding 3.6kN. This would not require a load cell to demonstrate if the method was followed correctly. Using 3 strong people on the same 3:1 (2.5 actual) system would probably be reasonable too.

50kg x 4 people = 200kg x 2.5 mechanical advantage = 500kg or 5kN
50kg x 3 people = 150kg x 2.5 mechanical advantage = 375kg or 3.75kN

Conclusions

Using a system like the one shown here, with 4 people pulling at average strengths, you can apply a force greater than 3.6kN to your test belt.

Once the test is complete you should thoroughly examine the belt like any other item of textile PPE to see if any damage or slippage has occurred. Any that do show signs of damage should be retired. Any slippage may be down to the buckle, but if the belt comes off or strap slides through the buckle under load, it should be deemed as having failed. If a belt has taken the test load and shows no damage or deformity then you can be comfortably sure that the belt will be fit for its intended use whilst still in that condition.

Final inspection of belts:
Lyon roller buckle    5.9kN    No damage
Caving Supplies square buckle 7.74kN    No damage
AV maillon closed harness buckle 7.7kN    No damage
Warmbac square buckle    8.64kN    No damage, slight curvature to webbing now when hung vertically which indicates over stretching or broken fibres down one side.

Again, this level of force was beyond what you would test to, but demonstrates that the 4 person 3:1 pull will not damage a belt that is not already fit for the bin.

A Note on Load Testing PPE

We don’t load test PPE. PPE is supplied with declarations of conformities and CE/EN markings. So long as you purchase via a reputable retailer or from the maker, this is the evidence that the product meets the minimum criteria set out in its approval standard.
Caving belts are not PPE and have no categorisation under the PPE Regulations. It therefore falls to the user to ensure they are fit for purpose, and that may involve a test of strength as outlined in this blog post. Ultimately, you must conduct your own risk assessment and define a way to show they are fit for use, copying a blog post won’t cut it with HSE!

Inspections

Lastly – all of this testing and use is predicated on you treating your belts as an item of PPE. They should be purchased new, inspected prior to use and have a recorded inspection every 6 months like any other PPE item. They should be in the same good condition as any textile item of PPE and retired from service if damaged, worn, contaminated or subjected to any load exceeding their safe limit. It is recommended that anyone in charge of inspecting PPE be trained and certified to do so.

As a side note, I maintain that the Caving Supplies belts are the tanks of the heavy duty caving belt world and, if kept very clean, will ultimately outperform every other type or brand available. I think this test shows that well as the CS belt had at least 5 more years of abuse over the other belts. I will dispose of the Warmbac belt just in case but don’t tend to use these anyway, but that’s another blog post!

More force testing on 5:1 systems

Posted on 22nd March 2016 by Pete

This post follows up on some initial testing done on 5:1 mechanical advantage systems used to tension tyrolean crossings done a few months ago. I suggest anyone who has not read that report catch up with it here before reading on as I don’t explain everything again here.

For this batch of testing I used the same site but rigged things using metal strops instead of rope loops. This would act more like the solid bolt anchors used underground and would nearly eliminate false readings from knots tightening.

I used 2 types of readily available Type A rope

11mm Mammut Performance semi-static
10mm Beal Antipodes / Industrie

The tests were repeated with 3 different progress capture devices

Brand new Petzl Stop (rigged both fully and half threaded)
10 year old worn Petzl Stop fully rigged
Brand new Petzl RIG

I created a 5:1 system on 10m section of rope using a Petzl Ascension jammer, Petzl Tandem pulley and a Petzl Partner pulley. These are all items that would likely be used by leaders underground or of similar type. No big rescue pulleys or prussics.
I pulled all of the tests on my own with un-gloved hands. I weight approx. 90kg and pulled as hard as I could using just hand grip.
The final tension in the line was estimated by hanging off it and the force on the jammer ascertained using a Rock Exotica Enforcer load cell measuring in kN.

11mm rope

New Petzl Stop – fully rigged
2.06kN
2.10kN
New Petzl Stop – half rigged
1.96kN
2.04kN
Old Petzl Stop – fully rigged
2.04kN
2.04kN
Petzl RIG – belay mode
1.94kN
2.08kN

10mm Rope

New Petzl Stop – fully rigged
1.88kN
1.88kN
New Petzl Stop – half rigged
1.70kN
1.82kN
Old Petzl Stop – fully rigged
1.98kN
1.92kN
Petzl RIG – belay mode
1.78kN
1.80kN

There clearly was a drop off in force required to tension a 10mm system over the 11mm system, although only small. The fully rigged Petzl Stops required the highest force to tension although the old Stop in the 10mm test oddly required more than the new one (*see foot note).

I took the highest force generating configuration and added some more people to the pulling end.

11mm rope with a fully threaded brand new Petzl Stop

2 smaller adults pulling
2.00kN
2.22kN
2.34kN

2 small adults & myself pulling
3.56kN
3.24kN
3.54kN

I think it is entirely possible to exceed the 4kN figure if 3 large and/or strong adults were to be pulling on a 5:1 tensioning system. Both ropes used were clean and supple, with a dusty rope friction would again increase and coupled with some less efficient pulleys might tip the force higher still. I think that it is still appropriate to give out the advice that no more than 2 people are used to tension 5:1 systems, perhaps 3 if using youths or very small adults but certainly no more. The force required to damage a rope at the teeth of the jammer is rather large, especially on 11mm rope, but repeated tensioning on the same spot in the rope may, over time, lead to degredation of the rope.

The best advice I can give is to echo what is already taught at LCMLA and CIC:

Keep your pulling ratios at 5:1 or lower and don’t exceed 10 men equivalent pulling power. i.e. 3:1 with 3 pulling or 5:1 with 2 pulling.
Keep ropes clean and supple.
Use only Type A ropes compatible with your choice of progress capture device.
Thick ropes are stronger and stretch less but require more force to initially tension.
Thinner ropes are strong enough but stretch a little more and require less force to initially tension.
Where very high tension systems are required consider doubling up on ropes and using a non-toothed rope clamp like a prussic or Petzl Shunt / Rescuecender.

A final thought. It is only a short period that the tension is applied to the rope via the teeth of a jammer in these set ups. It is the resultant tension and forces in use that are just as, if not more important to keep an eye on. Tensions in tyroleans can easily exceed 2.00kN, the maximum load Petzl advise for a Stop descender. Consider all components carefully and practice safely before using for real.

* Having given this some thought I believe that I can explain the added friction for this configuration. Over its life, the older Stop has been used for many miles of 10mm rope, wearing the alloy bobbin into a matching profile. Now there is a larger contact area between the alloy and the rope when compared to the brand new Stop. The larger contact area requires more friction to overcome and hence the greater force required to pull the rope through.

Loads on a 5:1 Tensioning System

Posted on 14th December 2015 by Pete

Tyroleans have been a bit of hot topic with me recently. I’ve developed some sites to use in my woodland near Whaley Bridge and been involved in some testing with BCA Trainer Assessors for the LCMLA scheme. We’ve measured the actual forces held by the anchors in a number of tyroleans but a really interesting questions was yet to be answered definitively:

In using a high mechanical advantage tensioning system, how much force is being applied to the rope via the teeth of the jammer and could we be at risk of damaging the rope?

To explain, when using a 3:1 or 5:1 system as is common with tyrolean set-ups, a toothed jammer is most commonly used to create the attachment point on the rope to build the mechanical advantage system. The force applied by whomever is hauling in is multiplied in a mechanical advantage system, which is kind of the point, and all this force is transmitted to the rope via the toothed jammer. The picture below shows a 5:1 set up with a Rock Exotica Enforcer load cell.

If you omit the load cell from this set up you have a standard 5:1. As you can see it is the toothed cam on the Petzl Ascension device that is the contact point with the rope. This device, like many of the Petzl rope clamps, is approved for use with 8 to 13mm ropes but comes with the warning that the toothed cam can damage or cut the rope at forces around 4kN for smaller diameters and 6.5kN for the largest. As it is hard to compare one rope to another, even of the same diameter, most rope professionals simply take the 4kN figure as that which must never be achieved in use.

Using 2 people to tension the 5:1 system, the Enforcer gave a max force of 2.88kN through the jammer. Had we been on more solid ground (and my partner not been a positively tiny 5’2″ & 50kg) I think we could have gone higher.

Inspecting the rope (Gleistein 9mm Type A) after moving the jammer showed a flat spot and gaps in the sheath where the teeth had opened up the weave. There was some furring but it was impossible to say if this is new or was already present on this rope.

A repeat test on a different section of rope produced a force of 2.66kN and a similar flat spot and sheath opening.

We then set up a standard 3:1 ‘z-rig’ and repeated the test.

This test gave us a force of only 1.84kN using the same 2 person team with a less pronounced, but still visible, opening of the rope sheath bundles and overall flattening.

I think these observations uphold the understanding that the tensioning in tyrolean systems must be done with great care and by using the least amount of tensioning required for the crossing. I will conduct a further observational test at a real underground site with 10mm or above diametre rope for a comparison but the force figures will not be too dissimilar. It would be interesting to find the 2 heaviest/strongest volunteers I can and use them on a 5:1 system to see if it is possible to creep further toward the 4kN limit.

In conclusion, you can get close to, or potentially exceed, the 4kN safe load on a toothed-cam jammer when using tensioning systems in tyroleans. Tyroleans really are an element of verticality that you need to understand well and get training for to know how to be safe. Go and do a CIC/MIA/UKMR or other course or get in touch with me for a chat.

I’ll be investigating this further at some point but it might be worth looking at employing the use of a non-toothed rope grab like the Petzl Shunt or even an appropriate prussic knot as a way of limiting damage to ropes in high mechanical advantage systems.

NB – The current Petzl literature for the current Croll and Basic do not show a load at which the ascenders may damage the rope. These devices are sold as personal ascenders and are only labelled to take up to 140kg of user weight.

Woodland Tyrolean Development

Posted on 10th November 2015 by Pete

Recently I headed out to our private woodland site to have a play with my new Rock Exotica Enforcer. We have recently developed a tyrolean crossing here along with calculations of anticipated loads and safety factors. Using the Enforcer on this tyrolean would give a real world check of my calculated figures as well as giving me a relatively safe and controlled location to experiment.

The tyrolean spans a 20m gully and is rigged using large trees slung with Lyon steel strops and the tensioning is done on the lower end using a 3:1 system through a Petzl RIG clutch. In this testing I purposely tried to over tighten things to see how much tension, and hence force at the anchors, it was possible for 1 person to produce.
I used a 5:1 pulley system and installed the Enforcer between the anchor and the RIG so it gave a reading on the total tension force being held by the lower anchor strop.

The calculations I had done previously were based on an average weight of participant of 100Kg. The span was measured and the sag was estimated at 10% as in practice we’ve found it impossible to achieve less than that with semi-static rope (usually more like 15-20%). The load on each anchor (so x2 for the rope itself) was calculated using a number of methods, some involving scary trigonometry, but the simplest equation was:

Tension = (Load × Span) / (4 × Sag)
Tension = (100kg × 20m) / (4 × 2m)
Tension = 250kg (roughly translated to 2.5kN)

The WLL (working load limit, or safe working load) of each component was calculated at a fairly standard ratio of 5:1, that is a fifth of its MBS (minimum breaking strength). Using this ratio the lowest figure was 4.8kN for a Petzl OK Oval karabiner. Technically the Petzl RIG is weaker but as it will slip before it’s WLL is reached then it can be discounted*.

*providing the RIG is not locked off and the rope is dogged back into the rigging so a running slip could not result in a complete slackening of the system.

So the maths with a 10% sag gave me 2.5kN tension force on each end of the tyrolean.

The tension force graph is downloaded off the Enforcer to iPhone and then edited in Microsoft Excel looks like this, with Time on the x and Force in kN on the y axis. Click to expand:

The graph starts with me applying tension to the system and having a few test bounces. The main force peak near 2m20s is me hanging suspended and pulling myself to the centre of the crossing and bouncing. I then pull up to the higher end, take a breather, and run back off, giving the last spike.

The second graph was me installing the Enforcer at the high end of the tyrolean crossing and tensioning it back up again with the 5:1 system. The tension at the top anchor was a little less than before on the lower anchor but the peak tension (me doing a running jump crossing) was similar to the previous graph.

Some interesting observations from the day:

At no point was it possible to install more than 1.5kN of pre tension in the rope prior to crossing. This was tested up to 9:1 and on 2 different tyroleans.
Tension in the line always dropped after the first crossing and remained at or below 1kN. Probably after the knots tightened up.
Tension could then be raised back up with additional hauling but the force remained below 1.5kN. Do not keep re-tensioning in real use as the increase in tension for each loading may cumulate to break a rope. See BMC technical reports.
The peak force was close to our 2.5kN calculation. This is predicated on having at least 10% sag in the system. A set of specifications for mechanical advantage systems and number of people pulling should be set by a company to limit over tensioning.
Whatever the tensioning method, we could not achieve less than 10% sag on a tyrolean in use with a correctly installed clutch.
Even on a 9:1 tensioning system the peak force created was only 2.8kN, probably because the Petzl RIG slipped at that point.

The 2.8kN high figure was achieved on a 50m tyrolean set up later that day. Here I was using a 10mm Beal rope today with a MBS of 24kN and 5:1 WLL of 4.8kN. The 5:1 safety factor is acheived with this rope. The rope I have on order for this when it is done with the public is a Mammut 11mm, the same as we use on the 20m line. It has a MBS of 35kN and hence a 5:1 WLL of 7kN, far above the expected loads in use.

Why all this effort? I like to know that the real world forces are actually near to how we calculate things, especially in tyrolean systems. Vector forces are scary and you just need to watch YouTube clips of slackliners breaking rated kit to see that. I can sleep well knowing all our research and calculations are backed up by real world testing and I can be confident that we are delivering as safe a service as we can.

Thanks for reading and remember, I can be hired to come to your site and test your rigging with the Enforcer too. Contact me direct for a quote and a chat.

Pull Testing Ground Anchor Pins

Posted on 9th September 2015 by Pete

It is quite common that where you need to anchor ropes to things, there are no natural things there to anchor to (like boulders). Climbers and cavers get round this by installing bolts and/or ground anchors. A ground anchor is a (hopefully) long length of angle, pole or bar that is driven in deep and (hopefully) can hold a load due to it’s friction with the ground. These are most commonly seen at the top of moorland cliff faces, installed by rock climbers as a top anchor to abseil from or belay a second. As part of the process of looking into developing a mobile zip wire / team building exercise we had decided that some form of ground anchor would be needed. There are a number of options, from home made to high cost but we opted for the tried and tested ‘mooring pin’ approach as used by mountain and cave rescue teams.

Now, just before people get a bit worried, this is not for life preserving! These pins are intended to act as guy lines for a tripod structure in poor, rocky soil.

Myself and Beth hammered the 50cm long 16mm galvanised steel pins in, one to 1/2 way and one to 3/4. Both of these pins came loose with very little force, about twice average body weight. We then drove the pins in up to their necks and set up a 10:1 pulley system from the back of my van, Brick.
The pins held well but at about 300kg they bent and pulled out. Even rigging a pair in a V made little difference, in fact I was dismayed to see my £20 pins bent after about 5 minutes of testing.

The final test involved setting the pins up as done by MR and CRO teams. We needed a 3rd pin so I used a section of what looked like 14mm rebar fencing spike. The 3 pins were hammered in full length in a direct line to the load. A rope was then used to tie each in line, stacking the pins with clove hitches seemed easiest. The force was applied to the front pin which, as it moved forward to take the strain, came tight on the 2nd pin which in turn came tight on the 3rd. The 3 pins worked as one and we were able to pull a force of 4.84kN with no failure of the anchor.
We stopped at that as we were now in danger of breaking the smaller pulleys used in the MA system and my van was beginning to move.

In conclusion:
Good news – the 3 pin method is very strong and would make a suitable ground anchor for guy lines but also, if backed up to a second system, would be okay for rigging safety lines from. Also, fencing spikes are seriously strong!
Bad news – £20 worth of 12 hour old pins are now bent like beech twigs. These just were not up to the job, poor steel. I’ll be picking up some fencing spikes and using them in the future.

Pete Knight's Blog

Underground & Technical Ramblings

Tag Archives: enforcer