Why accuracy means more than just figures on a spec sheet
Accuracy and repeatability of part size have always been important. The big change in the last decade or so is that producers want to shift the attainment of this accuracy as far up the line as possible, to the cutting stage. This can often remove the need for subsequent processing. Cheaper components is the result
Putting more accuracy into the cutting process is a trend that is likely to continue. However the costs must be balanced against the benefit that greater and greater accuracy will deliver. These benefits will peak at a certain point and thereafter, for every penny spent on attaining more accuracy, the return on that investment will reduce. So how far should fabricators go in the pursuit of the micron? The first step is to analyze customers' real needs. When assessing the accuracy required of a certain process, the temptation is often to pick the most difficult part as a benchmark. But this may actually represent less than 10% of production, and trying to achieve it overall can result in an over-engineered expensive system that is wasted on the vast bulk of production.
In many cases the part can be re-specified, processed by other methods or subcontracted. A small section of the workload should never put an unreasonable skew on the overall investment decision as to which technology is appropriate. Splitting the needs down will show that a range of accuracies is required. The next step is to decide how to meet them.
But first weigh up the "real needs" of the product; stories are rife of clearance holes on 16mm bolts specified to ±.05mm. In some cases there are sound reasons for this. But in most, these tolerances are far beyond what is really necessary. We can compare the ultimate accuracy figures of different cutting technologies on paper, but these can bear little relation to the actual precision obtained in practice. The disturbing fact is that some users are paying for those extremes of accuracy while in reality they may be unable to take advantage of them because of external influences. Often the limitations of the manufacturing process make it actually impossible to achieve accuracies greater than 0.1mm. In fact, one of the biggest influences on accuracy of the parts from profiling machines is not the machine or the process but the material from which the parts are cut.
Material considerations For instance, on large parts it is not uncommon for material to move, visibly, while being cut. The natural tendency is to blame the heat generated by the cutting process. But it is often the case that this is the effect of the relief of stresses locked in by deformation that naturally occurs during the rolling process. No matter how that material is cut, regardless of the precision of the machine, that movement due to stress relief will still take place. In some circumstance, the movement could be as much as 1mm/m. A further problem is thermal stability of the material. If one part is cut on a Monday morning at 14°C and the mating part is cut on Wednesday afternoon at 21°C then, on a 1.5m long part, the mismatch could be 0.15mm. Yet this temperature range is quite common in practical manufacturing. Indeed, three plates stacked on top of each other in the sun would exhibit such a temperature range.
Lack of flatness compounds the thermal and stress relief effects - particularly in coil steel stock. A 6mm bow in a 1m long part produces a 0.08mm error in the length when the part is flattened. On top of these factors there are other influences due to variations in the process itself. Wear of nozzles, incorrect cutting height or speed, and inappropriate gas flow etc will all have an effect. Even with a laser these variables can easily result in a 0.08mm in variation in cut width. With all these "external" influences on accuracy, users should guard against putting all faith - and particularly money - in to the technology chosen to cut it.
It is true to say that the more accurate the cutting system, the more accurate, overall, the parts will be. But if you are using a technology that will cost say $1.00 per meter to cut the parts to an accuracy that is either not possible or at least unwarranted, and another process is achieving say 90-95% of the result at 50 cents a meter... then the viability of the $1.00 per meter process is seriously in jeopardy. Good advice is to take an overview of the job in question.
Don't slavishly follow the drawings
Don't look solely at the drawings, but look at what the parts are being used for, the subsequent processes they have to go through and talk it through with the customer. Try the parts using various technologies and submit them to the process. If all are acceptable in practice, then the process can be selected that gives the best economic return. Accuracy is not just a passing fashion. But there are elements of hype. Be careful not to let the glamour of accuracy cloud common sense judgments about all the other facets of the production cost cycle
