prevention of failure II – ribs – a designer’s perspective

 Variation of wall thickness at rib.

For those of you who have been following our recent rants and ramblings in our articles about the prediction of failures of tanks and chambers, you will probably be saying these guys just sit there steaming away in their digital world thinking about things which may, or may not, be happening out there in the real world.

Your right! We do spend most of our time hiding behind our monitors, in cyberspace playing with the bugs in our software, twiddling our thumbs and waiting for inspiration to arrive. But, sometimes we have interesting days when we are asked to visit a moulder, or a test house, with our client especially when a tank is being moulded for the first time or when products are under test to observe and advise.

On a recent entertaining day we see operatives swinging on the tool, hitting the moulded product with hammers or prising it out of the tool with crowbars. Once these minor physical issues are resolved and the operative made to see the error of their ways, the recurring question is, has the polymer been subjected to the optimal heating and cooling parameters during the moulding process?

Underground tank – thinning on tank riser pipe due to the incorrect size of radii – note this tank claims to withstand vehicular and pedestrian loading and is thinnest where the most strength is required!

The ribs we design on the circumference of a tank or chamber normally shrink away from the tool during the cooling phase of the cycle but ribs and geometric features at other areas of the tank may shrink onto the tool and sometimes limit the freedom of the main body of the tank to shrink and inspection and measurement of moulded sample products. We note the shielding fixed to specific areas of the mould and strategically located air movers, directing heat and intended to ensure consistent wall thickness at geometrically difficult locations.

The correct grade of compounded polymer is selected, flow tested for quality and carefully weighed out. The mould is loaded and closed for the moulding cycle. At the end of the cycle, the mould is opened and the first off is born and delivered into the arms of the waiting assemblage. Once the tank has cooled down, out come the white coats with their digital calipers to verify the dimensions and their ultrasonic thickness tester to check to be sure the wall thickness is within the tolerance quoted on the drawings; specifically at areas of high stress indicated within our FE analysis report plots.

Underground tank – daylight shining through thinnest section of solid rib and notches.

And then the operatives review the plots from the data logger, which confirm to us that the internal air temperature reached its magical peak for the said time. Samples are taken to the lab for impact testing, verification of density and subjected to microscopy to identify bubbles and inclusions, etc.

Okay we won’t go any further; it doesn’t happen in most moulding shops. So it isn’t surprising that we are increasingly approached by companies who have had failures with tanks and chambers. Handling, storage, distribution, installation, commissioning, servicing, normal use, abnormal use and misuse never contribute to failure of the tank or chamber do they? So let’s deal with manufacturing issues first!

It is generally accepted that in comparison to injection or blow moulding the rotational method is difficult to control since the variability inherent within the process may give rise to polymer degradation and/or poor wall thickness the shot weight defined and the tool has been manufactured and fitted to the moulding machine it comes down to process capability and control.

Underground tank – failure through the thickest section of a solid rib.

The age and type of machine together with operator knowledge and experience are paramount to the production process. We assume that the designer has done his job correctly and the toolmaker has manufactured the tool to be fit for purpose. So how do we assist the operatives to produce consistent product.

Readers will recall that in our previous article we discussed the pros and cons of solid versus open ribs and their relative merits in the design of tanks and chambers. The distribution. Combined with this are a whole host of other potential issues which singularly or in combination impact upon the structural integrity of the product and whittle away at the factor safety built into the product at the design stage.

Coincidently, with regard to factor of safety, we recently came across an underground tank with a wall thickness, which in our experience is much less than would be expected for the application intended. In the product literature the issue of whether it could withstand groundwater was defined clearly as “occasionally”.

Readers who saw our previous article on the failure of tanks subjected to groundwater will now appreciate that this statement is sheer nonsense since it implies that from time to time the tank may fail and that this is somehow acceptable. The term “factor of safety” seems to be the world’s best kept secret in some design houses and I hope MacDonnell Douglas or Airbus don’t let these lunatics loose in their design departments!

Once the product is designed, the materials selected, proportions of ribs and their relationship to wall thickness is as critical in terms of design intent and structural integrity as it is to process viability and variability.

The images indicate the variance achieved by one moulder who had little control of the moulding process. Sometimes the designer is forced to push the design and process envelope to meet strict functional and spatial limitations. In this case the moulder chosen experienced processing difficulties and decided that in lieu of sorting the problem to point the finger of guilt at the poor product designer and when that failed, at the toolmaker. The solution to the problem was simple. We arranged for the tools to be shipped to a competent moulder and the problem was solved in no time. In hindsight, we should have had the courtesy to buy the original moulder a mirror into which he could reflect. Horses for courses as the old saying goes!

One lesson to be learned is that variance in quality of the product is often a result of human intervention since there are an infinite amount of variables, which can be adjusted. Our simplistic view is to vary one variable at a time; a controlled approach to problem solving, which seems to be a rare attribute these days.

The use of temperature monitoring via a data logging is growing but the challenge it seems now is to come up with a simpler, cost effective and more robust system to win over the disbelievers. The heating and cooling phases of the Rotational Moulding process ensure that the internal air temperature of the mould is in a constant dynamic state of change. There are many papers on the subject and clear evidence that the impact strength of the moulded product is adversely affected by the formation of a degraded polymer skin on the inside surface of the product.

Illustration showing common faults with solid ribs

The issue of control is even more important in our experience in designs of tank which have solid ribs.

Rota Design works on an international basis designing, analysing and verifying designs using FE analysis techniques and procedures for clients in many parts of the world. During the design stage we are conscious that a design feature such as a solid rib may be required to be moulded at different machines within a facility or even in different countries and maybe using slightly different polymer grades due to local supply. This is one reason why we are so vocal about having a worldwide harmonised standard! We have found that each client has it own preferred proportions for solid ribs which seems to be based on a combination of factors founded predominantly upon local experience and personal preference.

Solid ribs for a new tank design, for example, may require that a R&D test tool is fabricated to establish the optimum geometric proportions for a given grade of polymer relative to the curvature of the tank and its nominal wall thickness. The test tool could include a number of variants of the proportions, for instance varying depth and radii, etc. The resulting samples produced may be cut through to inspect the cross section for voids, localised thinning and notches. The procedure to verify the optimum cross section for the solid rib is time consuming and may provide the optimum solution for the moulder but become problematic to produce consistently at multiple locations. A simple solution from our perspective is to recommend open rib configurations as the shot weight penalty is easily offset by the lower skill level required to mould the tank and the potential advantageous control of quality.

Another concern regarding solid ribs is based on our experience of the failure of a large chamber designed by a third party where the failure line ran along the path of the solid rib. Upon close inspection of the failure we concluded from the direction of the striations, etc. that it was a sudden brittle failure caused as a result of internal stresses building up due to the variable rates of cooling through the rib. So a geometric feature specifically designed to enhance the structural performance of the tank actually caused its demise! In this case we called upon our forensic procedure to
replicate the failure mode with in our FE software and then used these data as the benchmark for comparison and improvement.

Good design, coupled with FE analysis and design verification are prerequisites for the initial stages. The toolmaker has to work his magic and not compromise the design intent. The skill and experience of the operatives at the moulding machine are critical but to ensure that tank failures do not result out of poor processing, process control and monitoring is the vital element missing in many establishments which ensure consistency and reduction of quality costs.

It is this temperature feedback loop which is missing hence we see the variance in wall thickness on open ribs and the notches in solid ribs, all of which serve to eat away at the factor of safety specified for the design which allow it to withstand its normal service conditions for 20 or 30 years.