Lyle,

Thank you for the reference. Please note, the article has been rewritten by CKIT. Some details and figures have been altered/misplaced from the original text. My spelling is often not up to par, and the same can be said for CKIT's.

I applaud their effort, and if not for their diligence many of the original texts on conveyor technology would go unread.

I ask you to read the article and forgive us our transgressions. If you need more clarity on Belt Safety Factor there are other artices I can provide. The subject probably should have a dedicated paper on its substance. I did one for SME many years ago that has more substance to it. I did another article on steel cord splice design and their structural mechanics for BSH that also has more substance. ■

There is no real "rationale" behind these numbers.

Manufacturers used to sell their belts according to a working tension. But they were producing a minimum breaking strength.

So they had to "guess" what would be a "safe" safety factor according to the reinforcement they were using.

Cotton belt were given a 20 safety factor because of creeping and rotting.

Then came rayon(12), then Polyamid (10), then polyester(8). Each time they lowered the safety factor.

As for steel cord the strange figure of 6.7 is just two third of 10.

No magic behind that. Just, may be, experience, even if some so-called "experts" try to create sophisticated calculation to reach the same figure.

More recently, the safety factor used on long conveyors has been lowered (4.5/5) thanks to the use of very slow starts which brings Cd/Cn next to one. Then the allowance for the start can be removed. But splices must be done carefully... ■

Dear Jean Paul,

There is no real rationale to the selection of the belt safety factor? Hmmm?

Your dismissive arguments about the selection of the belt safety factor (SF) diminishes the extensive theoretical and laboratory test work undertaken by Hannover University, Rheinbraun Coal, and significant research by manufacturers and engineering companies in many places of the world.

Hannover University was requested, by the German mining companies, to develop means that defined the attributes which produced the necessary breaking strength, based on all known factors involved. This effort was undertaken because of the very significant underground fabric and steel cord splice failures of the 1960’s.and early 1970’s.

Hannover’s designed and built two research splice dynamic fatigue test machines, one for fabric and one for steel cord belt construction. In addition, ContiTech and DBT built identical machines based on the designs of Hannover’s Dr. Flebbe now with ContiTech..

These results then became the DIN 22101 standard “r” classifications for:

a) splice dynamics fatigue efficiency then = 0.359 (strength loss = 1/2.79 in one week)

b) degradation from age, factory and field installation error factors, et al = 1.00

c) acceleration and deceleration momentary factors = .0.400

d) running or working load = 1.00 = (basis for selecting the belt strength multiple)

Therefore, the steel cord SF = (running + starting + degradation) x (1/splice efficiency)

Based on this premise SF = (1.00 + 0.40 + 1.0) x 2.79 = 6.7:1 for steel cord

I do not wish to debate these above comments. I simply offer to the readers of this forum a known alternative and rationale point of view. This is based on my knowledge gained working with the original developers and so-called experts of DIN 22101. ■

An extension of the aforementioned SF derivation and today's lowering of the value to SF = 4.5-5.0:1 can be understood from the following.

-------------------------------------------------------------------------------

Modern techniques are better able to design belt splice fatigue strength to better than 50%-60% efficiency depending on:

a) Splice step design - step lap lengths with variable overlaps,

b) Core rubber fatigue properties- many mfgrs are superior today ,

c) Wire rope properties - understanding the elongations importance

d) Reinforcement methods - fabric layers above and below cords

About 10 years ago we designed and built a 2-step ST-5100 N/mm that achieved better than 60% splice efficiency. It was built, installed and work with success. This alone will yield a SF =< 5.0 : 1 ~ ( 1.0 + 0.4 + 1.0 )/ 050 (<0.60). The procedure is more complex than this simple analogy when we consider the million plus cycles the belt must endure as compared with the 10,000 cycles on the Hannover machine.

Starting and stopping factor 40% multiple over the running tension was developed to compensate for fixed-fill fluid couplings and for high initial step resistance of wound rotor motors. Today starting controls impose a much smaller factor. We derived the momentary starting fatigue properties based on sequential daily overloads placed on the same splice for 10x per day. When this factor was integrated into the overall cyclic factors its influence was reduced from 40% to 15% even with the 1.40 x fluid coupling start factor.

Elongation stresses, degradation from ageing, splice anomaly error and factor errors were set by DIN 22101 at a value equal to the running tension. Today we can refine factory errors, field construction errors, improve or eliminate bending of belts around pulleys, et al.

All of the above can be quantified with the original DIN 22101 numbers. Based on the analytic improvements, engineers can determine the degree of safety which meets prudent and legitimate design practices.

No black magic. No wait and see if it fails to put a point on the curve of experience.

We do begin to understand the successful low SF values applied to the Cable Belt designs. Cable Belt have typically worked with SF <= 4 : 1. The normal belt conveyor can endure numbers that compete with the Cable Belt concept.

I offer this to the interested reader without debate. ■

Mr Nordell,

All you write is right.

How can I dare to write in your forum?

I apologize.

Please forgive me.

I was just suggesting that conveyor belt design do include approximations because tests performed in a laboratory can not replicate the real life environment of conveyors.

But of course, I am wrong. You know better.

I have been an exec at the organization you mention for 17 years,

but I can't have the depth of knowledge you have.

I am really impressed by your supreme knowledge of the conveyor belt techniques.

It's a blessing to read you. ■

May I add other DIN 22101 omissions:

I offer an example of a short belt (600m long x 130 m lift) and a long overland (O/L with 10,000m long and 80 m lift). These two belts use the same ST-2500 N/mm belt strength transporting 3300 t/h of iron ore.

1. The DIN splice SF criterion does not consider short verses long conveyors. Thus, the number of high stress cycles is irrelevant between a 600m high incline short belt and 10000m long overland (O/L). Assume the short belt operates at 4 m/s and the long belt operates at 7 m/s. Typical numbers with today’s in-plant and O/Ls. The short belt will subject the splice to 9.3 times the peak stress cycles as the O/L, given the shipped belt rolls have the same length for short and long conveyors. However, we know they do not.

2. O/Ls typically have a thinner top cover per the manufacturer's same belt life warranty. This could be a 5-7mm difference, which means larger meters/roll and fewer splices per belt. With the short belt, at a top and bottom cover of 12 x 7mm, and the O/L covers at 7 x 5mm for the same ST-2500N/mm. This equates to 870m/roll on the O/L and 615 m/roll of the short incline (i.e. 870/615 = 1.41 times longer for the O/L). This means we need to make a correction to Item 1 above. The short belt will now have 13 times the peak splice loadings verses the O/L with this second correction ( 9.3 x 1.41 = 13.1)

3. The same splice is not positioned at the high tension location to subject to this peak stress for every high load incident during starting and stopping. If you have 20 splices, then each may take 1/20th the peak load incidences. There is no correction for this condition in the SF. For the O/L, only one splice will see peak load during starting, (ie. the belt will travel about one roll length during the starting of the 10 km O/L).

4. Belt conveyors are not normally operated at peak design conditions for which the belt rating is matched. Typically, the nominal operating load is only 70-80% of peak design. Thus, the fatigue damage analysis should factor this into the general SF.

When we consider all these factors, O/l's can have the Safety Factor (SF) reduced significantly.

I offer two antidotes:

1. We provided a big coal company a means to quantify the short and overland conveyor belt SF. Applying Miner's fatigue repetition rule, we determined short incline plant conveyors would need a SF ~ 10:1 for a steel cord construction with high lift. We further determined the proposed 10 km overland would need SF ~ 4:1. So, the client applied the SF~ 4:1 to the in-plant. The belts lasted less than 2 years. To this day they condemn the report.

2. A very high strength belt in Chile was designed with a SF~5.0 for starting and stopping. This is a large downhill. Big brakes are required. An engineer noted that brake pad wear can be reduced with faster and harsher braking. So they modified the stopping algorithm to stop ~ twice as fast. This reduced the SF ~ 3.8:1. The conveyor was about 5 km long. It has taken 100's of the high brake actions at the low SF, without serious incident. Hopefully, the practice has stopped. ■

David,

I agree fully to what you say. And I agree also that the safety factor has been a subject of a lot of research.

However, the day to day selection of a belt and the associated safety factor, is not an exact science.

I have been involved at the beginning of my career in the design of missile seekers.: we are very far from the precision required there. The tolerances in the belting business, from design to manufacture, make it impossible to assess precisely the parameters.

This is not to undermine the quality of the research. It's just a fact in an industry selling its products at roughly 4$ a kilo. ■