This article was first published in 2009.
I am not an engineer: I have no formal qualifications in engineering. However, over the last twelve years or so, my job has been in writing about technical – and often engineering – matters. Very often, that has meant designing and building what are frequently unique projects – mechanical, electronic, pneumatic and so on.
In my home library I have lots of books on engineering, and I’ve read and understood at least the broad thrust of most of them. But it’s always seemed to me that engineering books ignore what is surely the most important aspect of engineering real-world solutions to problems – and that’s devising new and unique solutions.
The books are full of theory, full of solved problems, full of examples where the solutions are just a few equations away. But unfortunately, in the real world – and especially for amateur engineers – these approaches are often a dead end. That is definitely not to say that the books are of no use – emphatically, far from it. But it is to say that, if you’re confronted with an engineering problem, the solutions may initially come from your head – and not books.
Instead, let’s take a look at what approaches can provide immediate and sound benefit when you, as an amateur, need to devise solutions to engineering problems.
One of the impediments is devising solutions to engineering problems is that thought processes are often constrained along narrow lines. Engineering textbooks are terribly bad in this regard – for example, car suspension textbooks are just filled with the designs that are already widely used in cars. A car suspension textbook won’t mention railways or horse drawn wagons or bicycles. They won’t mention aircraft undercarriages, or even sports shoes.
But if you want to devise unique solutions, you need to think of the topic in the broadest possible way. Suspension, for example, is fundamentally about springing, damping and linkages. In more complex iterations, it’s also about interconnectivity eg of wheels. If you examine the topic as widely as you can, you’ll gain a fair better understanding of what is actually possible. Taking our suspension example further, here are some reasons to examine these other vehicles:
Railways (linkages, interconnectivity, lateral springing, steel springs, rubber springs, air springs, hydraulic damping).
Horse-drawn wagons (large diameter wheels, lateral and longitudinal springing, timber springs, steel leaf springs, friction damping, motion ratios).
Bicycles (motion ratios, air springs, steel springs, plastic
springs, hydraulic damping, friction damping, air damping, oil damping, linkages).
Aircraft undercarriages (interconnectivity, rubber springs, elastic cord, nitrogen springs, hydraulic damping, linkages).
Sports shoes (hysteretic
damping, plastic springs, air springs).
I am talking in this article about innovative design – so, in the case of suspension, not just for example changing the springs in your Ford. If all you are doing is the latter, this article won’t be relevant. But if you’re trying to devise a suspension that, for example, is ultra-lightweight, has long travel and uses hydraulic damping, you should first think laterally about every device that uses suspension.
And it’s not just suspension. If you’re trying to devise a new intake manifold to take advantage of tuned pulse behaviour, you need to look not just at car (and other) engines, but to also consider the way that sound waves behave. Here’s where a textbook is great – but not one on engine intakes, but one on sound waves, eg how they travel and reflect. Most ‘old fashioned’ textbooks on sound will start off with the fundamental ideas – ideas that you can constantly be thinking about in the context of engine intakes, not the musical instruments often used as examples.
Limiting yourself to pursuing knowledge only within the narrowly defined area of interest means that you immediately constrain yourself to taking the path that others have already taken: in other words, innovation will be lacking.
Identify and prioritise the elements of the problem
Any engineering solution for any problem will achieve some outcomes better than others – you can’t solve everything at once to the same degree. Therefore, break the design down and then prioritise what problems are most important.
In the real world, for amateur engineers, often ‘cost’ and ‘ability to build the design’ are very important limiting factors. If you can’t afford it, or can’t make it, then the solution is clearly a failure for you. But after these factors, what are the next most important?
At the time of writing, my current project is the front suspension design on a recumbent pedal trike. The priorities are:
Able to be built and afforded
Stiff in roll
Long suspension travel (100 – 150mm)
Sufficiently durable to cope with 40kg per wheel loads and 1g bump
Able to be used with steering that can achieve no bump steer and have Ackermann geometry
It was only when I realised that ‘stiff in roll’ was so high up the list that I started to radically change my design ideas.
If you’re engineering a bracket, what’s more important? Weight? Strength? Ease of making? Appearance? Corrosion-proofness? Or does the bracket simply have to be made with materials you have to hand, right now?
Breaking the problem down to its component parts and then prioritising criteria lets you see the trees in the forest.
Consider the Views of Others... but often, ignore them
Especially with the speed, reach and access of the Web, these days you can always find someone to discuss ideas with – no matter how esoteric the engineering field. That can be of great value – but more often than not, unless you apply a good filter, it can stop you succeeding.
Consider for a moment a group of people. In that group there’ll be only few who’ll be really smart. That’s a politically incorrect thing to say, but it’s true. Therefore, the feedback you get on any idea from such a group is quite likely to be not well thought out. It may even be misleading, wrong, stupid, slightly incorrect, ridiculous – and so on. In fact, if the group with whom you are communicating is typical, and the idea is new, the feedback is much more likely to be wrong than right.
And it doesn’t matter how many in the group say it: after all, by definition, if you’re coming up with a new idea or a new approach, they haven’t already thought of it. And, human nature being what it is, people tend to reject new ideas.
Over the years I have received some excellent advice from Web people I’ve never met. But you really have to cull out the vast majority who are simply not equipped to give you useful advice. The ‘old fashioned’ ideas like assessing the background of the person, finding out what they have achieved in the field you’re discussing, considering how lateral their thought processes are, and ascertaining whether they’ve ever actually carried out the idea (or a variation of it) - all must be employed if you’re to get advice that is of use.
That’s a very simple notion – don’t listen to idiots – but I see so many people who come up with a great idea, then retreat, tail between their legs, when other people dump on it from a great height.
Remember, also, that the more expert someone is at a topic, the more they tend to accept the prevailing wisdom. Therefore, ideas that challenge that wisdom are more likely to be rejected.
Let’s look at an example – you’ve decided that insulating the full length of the exhaust pipe will keep the gases hot and so make them flow faster, improving engine power.
You ask for some feedback from a group and people come back with:
“Won’t work, don’t waste your time.” (This person has contributed nothing; ignore them. Totally ignore them, don’t even let their comments penetrate your consciousness.)
“That approach was tried in the Sixties, it didn’t work.” (Nearly as pointless – no evidence, no detail. Maybe ask them for some detail - invariably, you won’t get it.)
“I tried lagging my extractors and I got only a 2 per cent gain on the dyno – so I decided it wasn’t worth it.” (Very useful – ask more questions.)
“I did this once and the drop in engine bay temp was great. Before I did the insulating, the engine bay was stinking hot; afterwards, it was clearly a lot cooler. Fuel economy seemed to go backwards though.” (Very useful – ask more questions.)
“The heating of the exhaust gases will cause greater refractive index, which in turn will cause the speed of sound to alter. This causes rarefactions of the exhaust gas that will lead to poorer flow. Or that’s what my old fluid dynamics lecturer told me, if I remember right.” (Absolutely typical of the rubbish I see all the time. Totally ignore them.)
“If that would work, why doesn’t everyone do it?” (This sort of response enrages me: often the answer is that no-one’s ever thought of doing it....)
Remember, if you’re trying to devise unique solutions to problems, by definition no one else has done the same before (or anyway, the solutions aren’t widely known). Therefore, don’t expect support for the ideas – you usually won’t get it.
A simple test is worth any textbook full of theory. Theories are simplified models of the world; in many cases, so simplified that they are relevant only when other factors are constrained to a silly degree.
Engineering mathematical analysis can show that a bridge won’t fall down. But no it can’t: the maths will tell you nothing about the homogeneity of the steel, the quality of the labourers, the presence of terrorists who next year destroy that bridge. Of course the maths can’t tell you that, you say. So, in fact, the maths cannot tell you whether the bridge will survive or not – it can tell you only about one aspect of its structural design.
The example may be a little silly (engineers don’t claim that the maths will tell you about subsequent terrorist attacks) but the principle remains sound: engineering theory is heavily constrained by the elements used in the model – include all elements that actually apply and it gets too hard.
That’s why real world testing is so good – it shows you what is actually occurring, not what is predicted to occur. If you are exploring a new solution to an engineering problem, do tests as early in the process as you can. Try also to do the tests in the simplest, cheapest way. Simple, cheap tests do not have to be poorly thought out or ineffectual. They may give uncertain answers (fine, test more thoroughly), they may give a really positive result (great - keep developing!) or they may give a really negative result (great - don’t waste any more time!).
People often seem terribly reluctant to do any testing, preferring instead to theorise. But because the test determines what actually happens, and theories are just simplified models of reality, if it’s at all possible, test, test, test.
So if you want to know what insulating the exhaust can do, it’s worth spending the money and lagging the bloody thing – not spending days or weeks in web discussions, reading thermodynamic textbooks, talking to learned professors, etc, etc. Just do it and find out!
In the case of my recumbent trike front suspension, after thinking about it for a while I built the design. The testing – riding over specific bumps, cornering on a skidpan against a stopwatch, data-logging vertical accelerations, taking action pics of full-load cornering – showed many things were occurring that I had never even considered. Some of those things were good, some of those things were bad – but now I know exactly what actually happens in the real world.
In some cases, of course, testing cannot occur of the full design – it’s too big or too expensive a project to build just to test. In that case, see if you can model the results. After consulting theory textbooks on the geometry of the required Watts Link for my trike suspension, I built a wooden model and tested it. The suspension at the rear used a space-frame that was modelled with the approach shown at Zero Cost Modelling of Space-Frames.
In electronics and pneumatics, it’s often possible to rig up a quick and dirty project and then test it to see if the results are as expected.
I once worked with an electronics engineer who had designed literally hundreds of projects, many of them unique. His recurring refrain: “Something unexpected nearly always happens when you test the prototype.”
Never mentally commit to an approach until there’s been a demonstration – of some sort – that it will work.
Using plastic sign material and gaffer tape, I trialled an undertray on the front of a Toyota Prius. It worked well, and I made the final version from engineering plastic. I installed another trial undertray on the back of a Honda Insight and it made no difference – so I took it off and the project proceeded no further. After the Prius results I’d been expecting the Honda undertray to work, but it didn’t and so I was glad that I hadn’t spent the time and money building a ‘proper’ undertray.
Let’s recapitulate the points made in this article.
Consider the views of others... but often, ignore them – seek advice and feedback, and bounce ideas off others. But closely analyse the answers, rather than just accepting them. Keep in mind that most innovations are initially rejected by a majority.
Testing – as soon as it is possible, test the solution. That could be in model form, or in a hastily rigged-up version. Remember, no theory is as good as a test of what actually happens.
Finally, I buy every engineering textbook on nearly every topic I can find. But when devising a turbo boost control system, I might look in my book on industrial pneumatics. When designing a recumbent trike frame, I might look at a book on airship frames of the 1930s. One of my favourite books a month or so ago shows line drawing after drawing of horse-drawn vehicle suspension systems. Why? Because I was wracking my brains to come up with something that would fulfil my front trike suspension criteria. Books are great – but you need to use them to enhance creativity, not stifle it.