Dear Gareth,

The equipment you point to in your attachment is called a Morstein twisted turnover. I suggest you not use it. For starts, the support rolls shown have high edge stresses that wear on the roller edges and crease the belt. Eventually, you inheret a 3 piece belt, unless the covers and reinforcement is sufficient. THe reason for the wear is differential speed between roll center and edge due to camber. They have failed belts as narrow as 1200mm and ST-4000 N/mm class.

REI tried making a three segmented roll assy to overcome this edge effect at Quintette coal in Canada in the mid 1980's. They still caused belt damage.

We have designed many ( maybe hundreds) turnovers some in the class you note.

There is more to engineering this size than you will find in any text. CDI can help as a consultant. We like you work for a living. CDI has invested significant time and expertise to develop these technologies. Thus, they come with a request to respect such efforts.

You also need to be very cognizant of the end supports where the twist is arrested and what is an appropriate twist length, not according to Phoenix. ■

Ref - attachment : At least in the Guided & Unsupported versions the belt centreline remains straight.

With the Supported Turnover gadget, as shown, the belt centreline travels along a helix over the barrel roller crowns. Shouldn't the contact points of the barrel crowns be in a straight line which line should be collinear to the entrant & exigent belt centreline? It is hard to prove that the inclination of the support tube is sufficient to maintain the required condition. There must be considerable lateral displacement, at least. It's a bit basic for belt training never mind the increased local edge tensions referred to by Nordell.

I don't think there was much checking of the source document, whether DIN or Phoenix, which is a bit of a shame. ■

Louis,

The crossectional shape of the Mordstein Turnover is tubular in shape. The roller placements induce the tubular configuration by applying a force offset to the crossectional centerline that will cause cupping across the belt width.

The purpose of the cupping is to shorten the path of the steel cord belt edge vs. its centerline length. The reduced path allows a shorter turnover length on the order of 14-17 x the belt width.

Standard flat ribbon turnovers require a turnover length of 25-35 x the belt width to maintain a reasonable edge safety factor.

CDI developed a better method of analysis, for standard helix, than is published in belt manufacturers' catalogs as noted by Adi Fritella's post on another thread. Ryan Lemmon was a CDI employee and granted the right to publish the work on its publication date. This method is included in the BELTSTAT conveyor design software. The main differences are the treatment of belt sag force (mass & width) along and helix path, transverse thrust, and use of intermediate support rolls such as at 45 degree twist position that reduces the peak sag force.

When conveyor belt's are larger than 1800 mm we resort to Finite Element Analysis (FEA). ■

Hi all..

I too am busy with a couple of wider (2100mm) steelcord overland conveyors with belt turn overs.

I am proposing to base the turn overs on a design from Sanvik where the belt is squashed into a flattened C shape in the middle, so it looks like a deep U shape that has rotated 90 degrees.

I find it a far more attractive solution than that shown in the Phoenix diagram with the internal pipe with loads of wheels on it.

Has anyone had experience of such turn overs? The roll arrangements seem quite simple.

Cheers

LSL Tekpro ■

I am very curious about the "C" section configuration.

We conduct FEA analysis on very wide belts and have never seen the visionary "C" section. I have seen where the center of the belt is forced up to collapse the outer edges inward and not use supporting rolls in the twist length. This was only used on small and light ST-rating belts. The twist length was very small and the belt could support itself over the span.

Mordstein wrote of the various types when he did his thesis on turnovers. The pipe support and twist rollers were his invention. There is an Achilles heal to this method. When the belt is cupped into a modest "C" the contact rollers do not have the same contact velocity across the supporting (Mordstein) rollers. Some have tried to overcome this anomaly by segmenting the support roller into a center and outer set of separately supported rollers. These too still cause excessive wear on the roller faces and where they crease the belt.

For wide heavy belts the sag stress and twist stress with the inherent bending thrust can be dominant with the flat ribbon helix twist. We use multi-point support to control the sag stress to an acceptable level. I do not know how your "deep - C" accomplishes this event. Can you draw a sketch?

Ever eager to learn more Graham. ■

I might add that I have seen a number of cases where the unsupported "C" section collapse near its center and not come right when it approaches the exit. The result is a mangled belt folded in half that does not open for loading as it exits the turnover zone. Big spill. ■

Hi Larry..

The turn over appears to be the Roxon belt twist design as used by Sandvik and I will e-mail the information about it to you.

Cheers

LSL Tekpro ■

Good-on-Yu Graham.

I see the point. Interesting. They took the natural shape produced when a flat belt is twisted 180 degrees. It cups at the mid point with the edges trying to minimize the cable stresses. They then applied a center roll to support it, at least for the bottom section.

I note your point to also support the top strand at the mid support. It will help a little to minimize the high bending-collapsing stress at the mid station.

I offer to assist you for a trade. No money. We can model the configuration, as it is similar to pipe conveyor technology. We do model turnovers now, just not this variety. The point will be that the internal torsional stress on the rubber-cable interface, at the center bend, will need to be kept below a critical threshold so it does not delaminate.

Hope to hear from you here or email. ■

Thanks for your offer of assistance Larry..

(And for free too!.. I must be doing something right at last)

Anyway, I am busy starting with a quick computer model of the horizontal "C" turn over arrangement to check for stresses in each cable and optimal idler configuration and position.

Should I find that (due to my excessively sheltered existance) I can't get decent results from my trusty Sinclair ZX 81, I will take you up on your kind offer.

Thanks Larry

Cheers

LSL Tekpro ■

Dear Graham,

I believe there is a better way. When you have done your buffin, or is it buffon routine, you can take a gander at another way.

Did anyone patent this concept? I must guess not or you would not? Anyway, I await your kind reply. ■

Dear Graham,

I also note the need to support the C section from the dirty side contact, if there are intermediate supporting rolls. This can create a buildup and related problems. If you try to use the ribbon, without intermediate support, you might find the sag force to be significant and excessive. We have experience with larger spans that are over 40x the belt width for wide and heavy belts where the total stress at the belt bottom edge is minimized. ■

Thanks Larry..

Busy with the calcs, and it looks good so far.

Loads and sag look fine. I'm using 15 x belt width for the turn-over, but have still to calculate the additional edge stress at this value.

I have the dirty side on the inside to contain spillage in the twist. This also means that it is the upper (inner) stub roll that is in contact with the dirty side. I am happy with this as it takes the lesser load.

Cheers

LSL Tekpro ■

Graham,

I would think one purpose of the turnover is to produce controlled spillage a one desired location. Not discharging the loose material either places some into the atmosphere as dust pollution or later discharge at the opposite end when you need to place the dirty side out. At the tail end, assuming incline transport, the opening will surely dump product on the ground after dusting the landscape.

I also observe, at the head return, that the C-section support rollers are best placed on the bottom outside (clean side), with possible support on the inside (dirty side) as you note. The inside upper support should likely be a rubber donut, high-low profile, to minimize build-up.

Only 2 cents worth. ■

Graham,

The upper inner support roll you reference has one difficult detail: the belt has a tendency to bend and apply significant extra pressure on the inner corner where the belt makes the C-shape. This may be a problem you do not want. Roll wear and belt wear can be problematic at this point. Add grinding paste = more grief. ■

Yes, We do this for a living. It can take a fair size model with composite structures, wire rope coupled torsional and tensile moduli, fabric orthotropic moduli, and rubber coupling to the above with non-linear behavior assembled to identify the submodel parameters, of sufficient size, then apply to major full size dimensional model.

You also need to have the model extend the boundary ends well beyond the pulley terminus to incorporate the end effects of multiples of the belt width. ■

Hi there Larry (et al)

I have now returned from site having tried to commission the belt turn-overs on two 2100mm wide steelcord overland belts using the "C" section central support system.

Suffice to say these turn-overs, or more accurately "roll-overs" are not successful, and quite frankly should be avoided at all costs. They are fiendishly difficult to set up, and have minds of their own when it comes to centralising the belt.

I reckon if I had enough time I could eventually get them to work, after a fashion, but they get a big zero from me for reliability and peace of mind for long term operation. They are a very high risk area.

I agree with Phoenix when they say "Only consider turn-overs when all other methods to clean the belt do not work" I also note they go on to say that "they should not normally be used for belts over 1400mm wide".

Come to think of it, I also question their real benefits. I simply have too many long overland conveyors, curved ones too, which are absolutely fine without turn-overs.

We have removed the roll-overs, and now all is well, and we can all relax.. forever!

I have learned enough from these roll-overs for a whole Beltcon paper. (At least I got rid of them before those annoying men in white coats came after me..)

Cheers

Taggart LSL Tekpro ■