Engineering design standards

[bernardgreen]

Domestic tumble driers that catch fire, cars that catch fire, second recall issued.

Other items with hazards but so far less publicity.

Who designs these dangerous items ? Where have good engineering skills gone ?

What has happened to the design process to alow these design errors to go through into full scale production and sale to the public.

 

[noseall]

My guess is that they are always on the limit in terms of quality v's safety v's cost and a lot of production models are actually test units.

 

[conny]

As stated, I think they are designed to the closest limits regarding quality, safety and cost. They are then only batch sample tested and quoted as being safe.
During production some number cruncher decides the profit margin is not good enough and says costs have to be cut or prices raised. Now no-one is going to want to pay more than the original price for the same piece of equipment so the manufacturer has to cost cut. They decide to get a cheaper thermostat or re-negotiate their present contract to buy more at a bigger discount, so this supplier has to reduce his costs to maintain his profit, and so it goes on down the supply chain. Quality then suffers which ultimately means safety suffers and things start going wrong.

 

[bernardgreen]

There is no doubt that cost to market has a strong influence on both quality of design and quality of assembly. But do design engineers ever point out that the cost reduced design has the potential to cause serious damage to equipment, homes and people. Should engineers point out the the consequence of poor design will damage the reputation of the manufacturer.

Maybe young designers are too keen to keep their job and thus will not risk pointing out to their employer that the restrictions imposed on them prevent them creating a reliably safe item. Fear of their (unscrupulous ) employer replacing them with someone who will keep silent and do as they are told. Maybe there are too many unscrupulous employers.

Or is it that design specialists can only see into their specialist area in the design and do not ( can not / not allowed to ) see the whole design.

And if they could see the whole design then could they detect the potential for failure or hazardous operation outside their area of speciality.

Is it that too many qualified designers are highly specialised and there are too few engineers who have general and wide knowledge ( and common sense ) that enables them to recognise where there is a potential for failure or hazardous operation. ?

Manufacturers like to employ fresh talent in the design department, fresh and lacking experience of the real world these designers can produce technically leading edge equipment that functions as per the specification ( most of the time ). But they fail to notice errors in the specification.

 

[securespark]

Many manufacturers use CAD and some designs are created by computer with very little being constructed. The Nissan Qashqai is an example of this.

Even if they go the traditional route and build working models etc.. they can be tested thousands of miles and still not show issues. It's only when you get significant numbers on the road that these reliability issues start rearing their ugly heads.

 

[MarkBarl]

I suspect that in the end it all comes down to cost. For example in the case of the wiring of a washing machine. Good engineering practice would be to size wires 1.5 times the expected current being carried. This makes the fault analysis easy as you can say you have applied sound engineering practice. If however you size the wire right on the limit, you will need to have detailed analysis to show that the conductor CSA is sufficient to meet the relevant directives to achieve your CE marking. It the manufacturer can save more in using smaller wires than the additional analysis costs they will go that way.
Additionally in manufacturers' failure modes and effects analysis normal practice is to look at the effect of a single failure at a time and not necessarily the cumulative effect. Therefore if something fails that would cause an increase current in a wire, it would not necessarily pick up that this might fatigue something else which in turn fails and potentially overload a wire causing it to catch fire. The relevant standards will make mention of ensuring that products are designed in such a way that there should not be cascading failures, however, these are often difficult to predict and designing to eliminate every eventuality is expensive.

 

[Himaginn]

Additionally in manufacturers' failure modes and effects analysis normal practice is to look at the effect of a single failure at a time and not necessarily the cumulative effect. Therefore if something fails that would cause an increase current in a wire, it would not necessarily pick up that this might fatigue something else which in turn fails and potentially overload a wire causing it to catch fire. The relevant standards will make mention of ensuring that products are designed in such a way that there should not be cascading failures, however, these are often difficult to predict and designing to eliminate every eventuality is expensive.

Systems Analysis, applied to failures, allows for and identifies cumulative, consecutive, cascading or simultaneous failures.
But, in all fairness, this type of systems analysis occurs after a failure, not before.

 

[MarkBarl]

Additionally in manufacturers' failure modes and effects analysis normal practice is to look at the effect of a single failure at a time and not necessarily the cumulative effect. Therefore if something fails that would cause an increase current in a wire, it would not necessarily pick up that this might fatigue something else which in turn fails and potentially overload a wire causing it to catch fire. The relevant standards will make mention of ensuring that products are designed in such a way that there should not be cascading failures, however, these are often difficult to predict and designing to eliminate every eventuality is expensive.

Systems Analysis, applied to failures, allows for and identifies cumulative, consecutive, cascading or simultaneous failures.
But, in all fairness, this type of systems analysis occurs after a failure, not before.

FMEA work should be used as part of the design process to identify potential areas of failure and mitigate a hazard. Unless there are specific standards that relate to a type of product, they would be required to put together a technical construction file which would include FMEA or FMCEA (Failure modes, causes and effects analysis) this should identify if one item failing would cause another to fail, but won't necessarily identify the effect of more than 1 component failing at the same time.

 

[bernardgreen]

Many manufacturers use CAD and some designs are created by computer with very little being constructed.

This can be a major cause of poor quality product design, the CAD system has been designed by a person skilled at designing a CAD package but with little knowledge ( for example ) about designing a car. Hence the CAD system's drawing tools and error checking tools may not be able to spot all the errors in the work of the car designer using the CAD system.

it is a bit like using a calculator, get the accurate result from the calculator but at the same time do a mental ( or pencil and paper ) approximate calculation to check the calculator's result is in the right range. Do the design on CAD to get the lines drawn accurately but then manually check that each line is in roughly the right place.

 

[mattylad]

And the person operating the CAD system often has not been fully trained on it so is not always able to use it correctly anyway. And when they do get the training they may not actually put it into practice for a long time so forget a lot.

CAD systems are brilliant and can make excellent products but it takes a trained person to use it correctly, and a trained & experienced person to use it to best effect.

Beancounters and managers make the products worse.

 

[conny]

Many years ago I worked for GEC, (is it permitted to state the company?), building distribution cabinets for industry.
Every few months a bespoke board would be designed and the customer would want as much as possible crammed into the smallest cabinet possible. The draughtsmen would do a rough design and then give the drawings and components to us to see if we could spot any possible future flaws. The most common flaw we usually found was if a component failed in the rear section, to gain access and repair it, (especially on site after installation), would require practically dismantling a major part of the board. Not a good financially viable option if it was simply a busbar bolt that hadn't been tightened correctly by the installer.
When younger designers started coming through and insisting the cabinets had to be built as designed we started losing business because they were also becoming more difficult to install.

 

[^woody^]

Inferior or faulty simple components from subcontractors or suppliers are often a cause.

But TBH, why after all these years no-one has perfected such basic appliances to run with no problems for years and years, is odd. Or do manufacturers need products to fail after a few years to keep sales up?

 

[conny]

If you developed something that never broke down then initially you would be on a winner.

However, after the majority had one your sales would plummet.