Jump to content







Photo - - - - -

Graphs And Charts In Engineering Design - Time To Say Goodbye To Them




Have you (today's young engineers) wondered how it was to do engineering design in the earlier decades i.e. the 60's, 70's, 80's of the twentieth century. The mainstay of engineering design then was an extensive usage of graphs, charts and nomograms for a variety of data such as physical properties, thermodynamic properties, friction factors, transport properties etc. These graphs, charts and nomograms served their purpose well in the absence of advanced computing machines in those days and some of the biggest chemical process plants existing today and running successfully are a testimony to their utility in engineering design done in those days.
I have myself used a lot of these in engineering design in the early part of my engineering career. The one thing that struck me about these figures and charts was that always there would be a small difference in the figure read from the chart or figure by two different individuals. Some of the petroleum related charts generated in the early 50's were so grainy and illegible that it became a herculean task to read numbers from them. However, those days we just had no option but to strain our eyes to read a number from the multiple lines drawn in a x-y chart. Many a times I would curse my luck if I had to refer to a old grainy chart related to the petroleum industry.

With the advent of advanced computing power through high speed computers and the introduction of regression analysis techniques using these high speed machines, charts and graphs are becoming obsolete today. Today using a fairly simple tool such as MS-Excel you can regress data to generate a variety of polynomial and logarithmic equations. Many enginering companies and academic institutions have taken up with earnest this task of data regression to generate mathematical equations which are simple to program and give accurate results for use in engineering design. Accuracy of the regressed data can be enhanced by increasing the number of polynomials in an equation. In these times when every minute you spend in engineering design is carefully scrutinized and costs need to be cut, these techniques are proving invaluable not only as a cost saving measure but also enhancing the accuracy of the engineering design.

As I have mentioned in my earlier blogs I am a person of the times. Leaving aside the initial hiccups I had from switching over to modern regression analysis techniques from the chart and graph era, I have now come to appreciate the usefulness of these modern methods and am now a whole hearted supporter of them since they have simplified my work and life.

I would love to hear from the veterans in engineering design their take on my views. And as far as the youngsters are concerned my advise would be to further develop techniques and methods to make engineering design more simpler and accurate.

Would appreciate comments and views from all forum members.

Regards,
Ankur.




Photo
Chris Haslego
Oct 29 2010 01:54 PM
I stil remember a few fold out charts from my engineering text books!
Your post reminded me that engineering is the product of many people's ideas & efforts to make it easier for future engineers to get the work done. Even the charts in the text book made me wonder how did people do it without (or with little) aid from computers/softwares!

Thanks for this post, it certainly made me more appreciative of the softwares so easily available today & I will be less grumpy with tedious software work! :lol:
As an "intermediate" process engineer (in my 30's) raised with diagrams but not using them so much I love the old curves and charts. Actually they should be posted on any engineer's office walls, as much as plant pictures and plant plot plans. They say and show a lot on the sensitivity of any phenomena. You can learn every day by just looking at a 3 phase diagram, or a Mollier diagram, or a set of chemical reaction kinetics curve.
I'm a second generation chemical engineer, so it's been drilled into me to look at everything from first principles. Thanks Dad. But then again I couldn't live without my Bonilla line sizing program, and would only refer to Crane if I need to do something new or different.

I like finding the balance between using old charts and computer programs. By using my McKetta Chart (GPSA) I can get a grounding in reality and then I'll use HYSYS for the rest.

Actually I'm a big fan of writing my own excel programs as it gives me a much better understanding of a problem and the assumptions that have been made. An example of this is when I was looking at the mathematics of pressure safety valves and that of flow orifices I found that they matched, but have different constants (fudge factors). Which is in a way first principles...
Charts and nomographs made you understand precision limits of wanted value visually (even in case they did not show scattering of experimental data). They promoted step by step calculation and physical insight, which can be lost today by e.g. merely running a simulator. Difficulty in reading them was sometimes intentional, not to provide greater precision than their own; but other times editors shrinked them to save space, causing trouble to readers.
Generally, graphical methods were widely used in Engineering before 1970 (e.g. McCabe-Thiele diagram for distillation, Cremona diagram for a steel bar network, bending moment / shear stress diagram in strength of materials), and are still used today as "exercises" to provide insight.
Main disadvantage of charts and nomographs is the fact you have to reintroduce a value every time you change a variable. Working in excell, I had to express a curve algebraically to avoid it. Of course algebraic expression is easier to use (but excell regression tools may not accurately express a few data, e.g. in the form y=(a+bx)/(c+dx)).
i hardly use any softwares for analyzing my chem plant, i still do a lot of paperwork like graphs and all... even now.
I totally agree with those above who have said that reading data off a chart gave engineers a better "feel" for the solution and its accuracy. Time pressures simply do not allow us to do all our calculations that way any more, but I still encourage young engineers when doing a calculation for the first time to use the charts to help develop this feeling for the job.
Its still useful for the engineers to stuck to the procedures for better in depth understanding, however you can verify with the softwares data bank the accuracy of the results.

Aziz
There are different types of Engineers.

One type would try to find a quick short-cut to every problem solving, i.e he would have used a chart or just know enough (from his peers) to be able to run a simulation program or a Company produced Excel spreadsheet to get to the result. These are the people who can quote a design philosophy to the last punctuation. To them Engineering is a means to the end i.e the result (irrespective of whether the result makes any sense or not!).

The other type of engineer is analytical and he enjoys the whys and wherefores of every problem. Whether it is a chart or a program, he does not just rest by putting the input and getting the result. He would look beyond the particular point on the chart and visualise the nature of the curve or the relationship with teach other on a family of curves. He is looking for interactions and meanings. His quest for knowledge is huge. He is into the Philosophy of Design rather than the very narrow 'Design Philosophy'.

With the loss of the Graphs and Charts, I am sorry to say that whatever was the opportunity for the first type of engineer to even notice or learn or visualise (even by accident) have gone. With project pressures he will churn out rubbish by putting in rubbish. He will never challenge his own result and the checker will be using the same tools to check his output. Hardly ever will the checker look out of the box to question anything. To both of them it is a unique number (derived from hallowed formulae or program) and hence the result must be good.

Even with the second type of engineer, his inquisitive and analytical mind will be stifled due to project pressures, and he will not have the opportunity of examining these charts and visualise the interactions and relationships of the other variables.

What I am trying to say is that it is not good that Graphs and Charts are gone or going, because A PICTURE IS BETTER THAN A THOUSAND WORDS. You do not have to use them exclusively but they will in most cases allow you to check back with the result produced by "unquestionable" algorithms. More so it allows the younger engineer to visualise the effect of the variables and gain the sensitivity of his chosen input variable.

I am sorry to say that many of the top level SIMULATION 'experts' I now see cannot even show on a piece of paper how an isenthalpic expansion swoops down towards the envelope, or what is indeed a Cricondentherm or a Cricondenbar. They cannot point out where is the dense phase on an envelope and visualise the isenthalpic or isentropic lines around that region. Many of them have forgotten where is the Critical Point in a multicomponent envelope and why it really has no physical meaning. They cannot qualitatively show on a hand drawn PT diagram as to how the envelope is modified when the original gas is dewpointed. Very few even remember the Mollier diagram and Phase Diagram for pure components and never took a journey by travelling along an isenthalpic or an isentropic line from say a P1&T1 to a new P2 by using a valve or a turbine. If you tell them that the path followed by a turbine will depend on its efficiency lying somewhere between the purely isentropic and purely isenthalpic drops, they will not know what the person is talking about. Yes, they WILL know when you start highlighting the two paths ON A GRAPH and showing them the difference. In most cases they come out very impressed in reinventing themselves as to their capability and new-found depth of understanding.

In this day and age of 'first oil' project pressures, it will be a very retrograde step to remove graphs and charts in Engineering Design.
The Type I engineer will become even dumber and the Type II engineer will have to work extra hard to visualise the interactions which were there for anybody to see and reconcile his calculated results.

No, I am sorry Ankur, removing the Graphs and Charts will only reduce the cementing of our theory with real life applications for the young graduates.

Latest Visitors

  • Photo
    sers
    Today, 11:58 AM
  • Photo
    Crows88
    Today, 06:13 AM
  • Photo
    DAJAY
    Today, 05:02 AM
  • Photo
    mudiah
    Today, 04:29 AM
  • Photo
    Shirley2012
    Today, 03:07 AM
  • Photo
    namita.modak1
    Yesterday, 10:35 PM
  • Photo
    Ganesh Kawade
    Yesterday, 09:57 PM
  • Photo
    pratnorris@gmail.com
    Yesterday, 04:04 PM
  • Photo
    amm2678
    Yesterday, 11:35 AM