Dimensioning a safety rail

[Walterke]

I've been asked to put a safety rail on a structure that has to be able to support 3 people falling off
The rail would be 4m long and the people are attached to 2m long safety lines.
So worst case scenario is all 3 of them fall simultaneously in the middle of the rail and fall down 2 meters before their lifeline pulls the rail.
If anyone has any ideas of how to calculate the required strength of the rail, your help would be appreciated.
(rail would be S355 steel and is allowed to deform but not break)

Google only gives me formulas where you have to 'assume' a certain amount of deflection to calculate the impact force, but that doesn't seem to be what I'm looking for.
I've tried using m.g.h=kx²/2 where:
k=48EI/L³
x=deflection of the rail
but kinda get stuck.

NX 7.5.5.4 with Teamcenter 8 on win7 64
Intel Xeon @3.2GHz
8GB RAM
Nvidia Quadro 2000

 

[desertfox]

Hi

Have you a sketch of the rail and how it is supported, also any details of the safety lines ie size, elastic properties etc
As a rough starting point I would take the average weight of a man (about 70kg) multiply it by three and then double it, giving about 420kg.
Add a good safety factor in to the figure above and go from there.

Have you looked to see what standards might cover this kind of equiptment or as your company done anything like this before that you could look at.

 

[Walterke]

The product gets sold to a company who then rents it out. Safety lines are to be provided by the end user so they can vary.
It's easy to calculate the required strength of the bar if I know the force that acts on it. I just don't know how to properly implement the "falling force" or whatever I should call it. (the effect of 3 men falling 2 meters down and that fall being stopped by the rail)
The rail would be welded into the construction (work platform), with some reinforcements at the end to prevent them from being ripped out.
Can't seem to find anything in the standard other than they have to be "strong enough"...


NX 7.5.5.4 with Teamcenter 8 on win7 64
Intel Xeon @3.2GHz
8GB RAM
Nvidia Quadro 2000

 

[LittleInch]

Someone falling 2 m then coming to a dead stop!! They will be talking in a high pitched voice for quite a long time.... You need to use a fall arrestor connection which slows people down and reduces your load considerably. The suppliers should be able to give you forces.

I've just done a quick first order calc and after 2m vertical fall, the velocity of your man after a 2m fall from a standing start is about 6.5 m/sec (Vel^2 = 2gh). Force to stop this in say 0.1 seconds is then F= M X A = 4900N - 490 kg force PER PERSON. This is about 7G loading so about right for a dead stop. Skydivers stop from about 60m/sec in about 3 seconds and hit peaks of 3G. I know from personal experience if you don't have your leg straps done up tightly it HURTS.

Therefore you need to allow for a force of about 3 tonnes to stop this failing in service. This is going to be some size rail...

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Walterke]

Contacted our customer and they use Shock absorbing rope lanyards featuring a built-in shock absorber which limits the fall arrest force in a fall situation to within 6kN.
So if I take 3x6kN=18kN and add some safety factor I guess I'll be fine.
You'll be pretty close with the 3tonnes :s

NX 7.5.5.4 with Teamcenter 8 on win7 64
Intel Xeon @3.2GHz
8GB RAM
Nvidia Quadro 2000

 

[Jboggs]

Everything in my gut tells me not to touch this one with a 10 foot pole. I've seen too many engineer's lives ruined by lawsuits from folks that misused, overloaded, or abused equipment they designed. That being said, if your boss is insisting on this, I would begin by contacting several safety lanyard manufacturers. They all have already done all the dynamic calculations and have published ratings of their lanyards. Because of the speeds and inertias involved, they are pretty high, in the thousands of pounds. Your rail should be able to absorb the full rated load of all the lanyards simultaneously. You should design it, build it, and load test it (both static and dynamic tests) at that load plus at least 25%. Get your lawyers involved UP FRONT so they can recommend protective measures for you (such as labeling, wavers, etc.).

 

[Walterke]

For the people interested: I did some more digging and apparently there's a European norm EN795 that says a falling person has to be calculated as a 1000kg weight.
Guess this'll give me enough to start with.

NX 7.5.5.4 with Teamcenter 8 on win7 64
Intel Xeon @3.2GHz
8GB RAM
Nvidia Quadro 2000

 

[LittleInch]

Like I said - 3 tonnes... They can't mean that each person should weigh 1000kg - even Germans aren't that heavy ;-)

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Walterke]

Actually since recently Mexico is, on average, the heaviest country in the world, followed closely by the Americans.


NX 7.5.5.4 with Teamcenter 8 on win7 64
Intel Xeon @3.2GHz
8GB RAM
Nvidia Quadro 2000

 

[LittleInch]

Amazing what you learn about - looks like 795D is what you want

http://hutter.onlineoplossing.nl/290/En795+norm+for+protection+against+fall+from+heights.html

Person of 100kg (a bit more like it) falling 2.5 m or static weight of 10000 kN (1000 kg)

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[racookpe1978]

Although OSHA only requires a handrail/guard rail be rated at 200 lbs force sideways, it has much, much higher rules for fall arrest systems and what they are attached to:

1926.502(d)

"Personal fall arrest systems." Personal fall arrest systems and their use shall comply with the provisions set forth below. Effective January 1, 1998, body belts are not acceptable as part of a personal fall arrest system. Note: The use of a body belt in a positioning device system is acceptable and is regulated under paragraph (e) of this section.

1926.502(d)(1)

Connectors shall be drop forged, pressed or formed steel, or made of equivalent materials.

1926.502(d)(2)

Connectors shall have a corrosion-resistant finish, and all surfaces and edges shall be smooth to prevent damage to interfacing parts of the system.

1926.502(d)(3)

Dee-rings and snaphooks shall have a minimum tensile strength of 5,000 pounds (22.2 kN).

1926.502(d)(4)

Dee-rings and snaphooks shall be proof-tested to a minimum tensile load of 3,600 pounds (16 kN) without cracking, breaking, or taking permanent deformation.

1926.502(d)(5)

Snaphooks shall be sized to be compatible with the member to which they are connected to prevent unintentional disengagement of the snaphook by depression of the snaphook keeper by the connected member, or shall be a locking type snaphook designed and used to prevent disengagement of the snaphook by the contact of the snaphook keeper by the connected member. Effective January 1, 1998, only locking type snaphooks shall be used.

1926.502(d)(6)

Unless the snaphook is a locking type and designed for the following connections, snaphooks shall not be engaged:

1926.502(d)(6)(i)

directly to webbing, rope or wire rope;

1926.502(d)(6)(ii)

to each other;

1926.502(d)(6)(iii)

to a dee-ring to which another snaphook or other connector is attached;

1926.502(d)(6)(iv)

to a horizontal lifeline; or

1926.502(d)(6)(v)

to any object which is incompatibly shaped or dimensioned in relation to the snaphook such that unintentional disengagement could occur by the connected object being able to depress the snaphook keeper and release itself.

1926.502(d)(7)

On suspended scaffolds or similar work platforms with horizontal lifelines which may become vertical lifelines, the devices used to connect to a horizontal lifeline shall be capable of locking in both directions on the lifeline.

1926.502(d)(8)

Horizontal lifelines shall be designed, installed, and used, under the supervision of a qualified person, as part of a complete personal fall arrest system, which maintains a safety factor of at least two.

1926.502(d)(9)

Lanyards and vertical lifelines shall have a minimum breaking strength of 5,000 pounds (22.2 kN).

1926.502(d)(10) 1926.502(d)(10)(i)

Except as provided in paragraph (d)(10)(ii) of this section, when vertical lifelines are used, each employee shall be attached to a separate lifeline.

 

[chicopee]

First you may want to consider separate life line for each worker if they are working on a swinging scaffold which from your OP seems to be. If so, secondly have the workers wear harnesses with shock absorbing lanyards attached to the lifeline with climbing clamps that would lock on the lifelines if these workers were to fall.

See my attachment that deals with the application of momentum, PE and KE for a falling object or person attached by a lanyard with and without shock absorbing material (could also be detachable stitches weaved into the webbing material). Note the column "Using manufacturers' reference material and since I had this sheet for several years thinking that it would never be resurrected you may want to research the numbers in those columns as they may have been revised. Millers fall protection equipment has a myriad of fall protection equipment that can assist you in your research.

 

[KENAT]

You could cross check your results with an energy calculation.

Equate the potential energy of the falling persons to the spring/deformation energy of the rail.

Although, without opening the link I believe that my be what chicopee is suggesting.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[Ron]

Not sure where you are located, but in the US, the Occupational Safety and Health Administration (OSHA) requires that each line be anchored to a reaction of 5000 lbf.

 

[chicopee]

The rail was considered a fixed item. My attachment was to determined the effects on a non deformable body attached to a lanyard. The objective was to ultimately determine if a human could die from such effect because a worker was killed in an articulated man lift which toppled over. This worker was not tied to the manlift basket and was catapulted to his death. The discussion centered whether or not he would have survived had he worn his lanyard with a safety belt (not a harness) knowing full well that the would have been banged up against the basket and incur internal injuries if he had.

 

[dhengr]

Walterke:
I would be very careful on this kind of design work. I’ve been involved in several law suits (not my designs) where the OEM had a fairly reasonablely designed piece of equipment which was then used incorrectly or loaded improperly or assembled improperly by some incompetents up on the roof. These people, or others, then came back after the OEM for damages caused by the users own incompetence. The OEM should have foreseen this possible misuse and designed to prevent them. There certainly is a need for this type of equipment when it is well designed and built, but the potential for misuse is pretty great, so be sure your liability insurance is paid up. You can hardly do enough in the way of instructions, labels on the equip., instructional videos, etc. to keep yourself out of court.

You say.... “The product gets sold to a company who then rents it out. Safety lines are to be provided by the end user so they can vary.” Which type of safety line causes worst loading and can all three happen at once and be applied at the same point? And you say..., “The rail would be welded into the construction (work platform), with some reinforcements at the end to prevent them from being ripped out.” Yes, it can yield and deflect, it just can’t fail during the incident, and is it the actual loading or the magic 5k? Who does this design to keep your whatcha callit out of the ringer? Who selects the lanyards fall arresting or not? Then, add to that the ambiguity of OSHA, et.al. and attorneys have a field day with this kind of b.s. I would expect each man to have his own lifeline and his own tie-off point. Then design these tie-off points for 5k at failure or pull out, and you have something reasonably defendable. I believe that if you can define the type of lanyard sued and the max. man plus equip. load you can finesse the 5k load a bit, with sufficient testing of that entire system.

 

[Jboggs]

As I said earlier, get your lawyers involved up front on this. One thing I learned when I worked for an insurance company that covered manufacturers of equipment is this: there is nothing you can do that will prevent your being sued if something happens. The only thing you can do is to take steps to minimize the actual payment you will have to make by showing proof of how you tried to anticipate all possibilities and took steps to prevent them.

And I'm not sure I would get down in the weeds trying to figure out the momentum and forces involved in a falling body. Strictly speaking your rail won't see those forces. It will only see the forces applied to it by the harnesses. And those harnesses all have maximum ratings. That should be the basis for your strength calculations.

 

[europipe]

Just did some work like You,
as EN795 sais the force for falling is 10kN for one person, for every person extra count 1kN (pt.4.3.4)because they never fall at the same time, there is always (a fraction of) a second between .

 

[Walterke]

there is nothing you can do that will prevent your being sued if something happens.

Fortunately we are not situated in America . People are slightly more reasonable over here.
I considered looking at the harnesses for the falling force, but found out the EN795 (european norm) is stricter, and since I don't know exactly which falling protection they'll be wearing, I'd rather just follow the norm.

@europipe
Thanks, I'm looking into ordering the norm as we speak( just got back from holiday). I will definitely look that part up since it reduced the required strength of the bar severely.

@racookpe1978:
there is indeed a big difference between hand rails and safety bars.
EN280 mentions 400N (~90pounds, which is ,surprisingly, quite a bit less than what you're saying)

@everyone suggesting I look up EN795: thanks, but I already figured that out quite some time ago (which you would've noticed had you read the thread).
@everyone else: thanks for your contributions!

NX 7.5.5.4 with Teamcenter 8 on win7 64
Intel Xeon @3.2GHz
8GB RAM
Nvidia Quadro 2000