Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

RE: Design Charts and Regulation

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Gil, and others

Design charts are important for rapid decision making, and they also provide
a reference for checking independent calculations.

The problem is when there appears to be a lack of independence, and over
reliance on those tables, without self-checking. 

For example, most engineers appear to use the Australian Steel Institute
design capacity tables for selecting hot rolled sections. Those tables
contain an explicit disclaimer, and warning to seek the services of a
qualified professional. Most engineers I have met think the cold-formed
steel structures code (AS1538/AS4600) is more complex than the steel
structures code (AS1250/AS4100). This is largely because they use design
capacity tables and have never actually done calculations to the steel
structures code (AS4100), only ever used the design capacity tables.

But look closely at those tables. Has far as I know those tables are
generated by a custom version of LIMSTEEL, created by the centre for
advanced structural engineering (CASE). LIMSTEEL is the main stand alone
software used for steel member sizing. LIMSTEEL is integrated into the two
main frame analysis programs MicroStran and SpaceGass. The only real check
on these tables, and software appears to be in the education system.

Then there is Queensland. The state with registered professional engineers:
far too many of which who think they can self-certify anything. South
Australia may not have registered engineers, but the legislation requires
independent technical checks. The result is that if the local building
departments get pedantic they reject: glulam tables and the likes produced
and certified in Queensland. Not a major problem, it generally means that a
builder, or timber estimator cannot use the tables, but a civil engineer
(B.Eng) can blindly accept the tables if they choose to do so. Blind
acceptance not a good idea given some of the deficient calculations I've
seen for other structural products. Though the Queensland branch of the
timber develop association is far more informative than the South Australian
branch, which I don't understand given supposed be part of a national
association.

Calculations however are only part of the issue. At the end of the day the
calculations have to reflect reality. In mechanical engineering I was taught
that finite element analysis/method (FEA) can be highly inaccurate, that its
value was simply to reduce the number of prototypes which needed to be made
and tested. Civil engineers seem to take FEA as accurate and increasingly
using for design of one-off structures with no calibration against a
physical prototype.

Whether numbers are plucked from manual calculations, software or design
charts, they have to reflect the real world.

Conducting the calculation is off little value if there is no independent
technical check on the validity of the calculation procedures. Thus three
test values do not form the basis of a reliable theory (eg. sinusoidal
curves for buckling cold formed sections). And 3 test values by CASE at the
university of Sydney, with no other research centre to back up and validate
is also not acceptable. There appears to be a clear lack of scientific
method in Australian engineering research. OK! Our codes are based on
research from around the world, but still the collective empirical evidence
is still relatively scarce scattering of points plotted against complex
curve forms.

Every product and building is a real world experiment. And the competence of
an engineer is not properly assessed until their design experiences its
design loads and behaves as expected. As such the competence of civil and
structural engineers largely goes untested, until its too late. Most other
engineered systems are tested immediately. It is thus important to have a
rigorous and robust control system for the design of civil and structural
engineering systems to avoid latent defects.

Assuming that a theory has been properly validated, then calculations based
on that theory for each and every project is a waste of time, and a waste of
paper, and waste of storage space. Of course is does generate need for
buildings, to store all the scrap paper. Calculating the same thing over and
over again also has the potential for generating and introducing errors
which have otherwise been removed in the past. Even with examinations and
licensing schemes there is still vast variation in the competence and
capabilities of practicing engineers. So if the capable has calculated we
don't want the less than capable recalculating and coming up with a
different answer: a defective answer. Tables, design curves, and maxima and
minima are all helpful towards staying in the ball park of an acceptable
design-solution.

The requirement is knowing the limitations of the structural model behind
the tables. For example in Australia few users of the residential light
timber framing code (AS1684), have knowledge of the structural model
presented in AS1684.1, and thus are not fully conversant as to when a design
is outside the scope of AS1684, and really requires engineering input.

The span tables however pose a problem. An answer can be obtained from the
span tables in a few seconds. Verifying that answer by calculation would
take significantly longer, adding extra load conditions and other features,
increases the time still further. The builders and owners don't understand
that a small change, stepping outside the scope of the prescriptive solution
can considerably increase the time required to produce an acceptable
design-solution and appropriate evidence-of-suitability.

Another Example. South Australia makes a big deal about reactive clay soils,
and typically requires an engineer to assess the footing. I say assess,
because someone else as already decided on a concrete slab on ground no
matter how inappropriate a solution such may be. The assessment is dependent
on a plasticity index for the soil, which is but a subjective judgement and
highly variable. The results of the soil heave calculation are then pushed
through one or both of two computer programs: SLOG or Chord. From what I
have seen both are relatively low quality programs: but there is no
alternative software. As far as I know the tables in the code (AS2870) are
generated by one of these two programs. As far as I can tell, what Slog and
Chord actually do is only vaguely understood by most users: hence no
alternative software, and very few engineers conducting the calculations
themselves. Most of the designs can actually be taken directly from the code
(AS2870), without the need for Slog or Chord analysis. But without such
involvement little else for civil engineers to do in SA: it seems some
engineers spend their entire careers pushing numbers through Chord and/or
Slog, and do little else. They contribute little of value to a project and
yet want the status of a Telford, Brunel, Navier or Coulomb. Actually the
builders like the Chord/Slog analysis because can typically get a smaller
footing size than in the code: saving concrete and trenching. Though a
better solution may have been not to have cut and filled the site in the
first place: and provide a suspended floor with usable space under.

And then there is RAPT and concrete design in general. In the past few years
there have been several tilt-up wall panel collapses during construction.
Why? There are design curves available to get in the ball park, so engineer
should have time  to check handling requirements. But then met engineers who
don't know we have standard connections manuals for steelwork, and carry out
about 1/5th of the calculations to justify smaller connection components. So
why would they know about other industry publications like those for
concrete: which they apparently spend most of their time designing: if you
call pushing numbers through Chord/Slog design. It seems far too many
engineers in SA think, putting more steel into concrete is the solution to a
problem: neglecting that the material is supposed to be reinforced concrete,
thus requiring concrete around the steel. Crunching the numbers is pointless
if do not get the qualitative aspects of the design correct.

A few years back my father did certification work, and apparently lots of
the concrete designers were pushing numbers through RAPT and other software
and clearly not reading the results, or have any understanding of the
results. So how does a software company evaluate the extent of feedback?
Little feedback and the software is perfect or just indicative of a lot of
incompetent users?

Or similarly for products. Precast concrete floor panels, single simply
supported span tables. Engineer using as part of continuous frame, but
relying on the tables, with no assessment of the continuity. Qualitative
understanding, not capability crunching numbers.

The IEAust this morning emailed policy on draft policy for national
regulatory framework for engineers. Creating legislation to limit practice
to engineers. Why? Licensing doesn't work. Queensland has registered
engineers: much of the damage during cyclone Larry was due to defective
design.

If the IEAust was more concerned about competence and public welfare than
the status and prestige of persons with a B.Eng. Then it would not start at
the top of the hierarchy it would start at the bottom. First recognise that
the IEAust does not recognise the entire engineering team: because the
industrial relations system defines technicians who have less than two years
education. So the team comprises of:

1) Technicians (1 year)
2) Engineering Associates (2 year)
3) Engineering Technologist (3 year)
4) Engineers (4 year)

This education is largely irrelevant to the practice of engineering. Many of
the 4 year qualified are operating in positions where the work can be
carried out by technician level designer/analysts. If the so called
engineers took the traditional route from shop floor, through tracer,
drafter, designer up to engineer, then they would know the capabilities of
the other members of the engineering team and utilise their skills more
efficiently. If this was to be done their would be no shortage of engineers.

If start at the top level: then who decides who is competent. For B.Eng
MIEAust CP.Eng NPER is clearly an unreliable indicator of competence. But if
start at the bottom, define the requirements to be a registered engineering
technician (design/analysis), and make it a required stepping stone. Doing
so makes a larger body of people available to carry out the task, but
removes those persons without adequate experience and education from the
area of activity. It does not result in over night their being a shortage of
people with a B.Eng and adequate experience. All technicians, associates,
technologists and engineers have the potential to register, and be required
to be registered. Then additional constraints on practice can be imposed
which then gives rise to create a higher level register for associates, and
so on until the engineers register is created. But all have to pass from one
level to the other, no one can be registered as an engineer until they have
served their time at the lower levels. So that when a technician passes a
project onto a higher level, they can have confidence that they are getting
the services to be expected of that higher level.

Currently I believe if I was to turn to the services of an engineer, all I
am likely to get is some rubbish pumped out of some computer software. I
have little confidence that I can turn for guidance from someone who is
truly qualified as an engineer. The B.Eng is worthless if the graduates do
not make use of the content. And the work most are carrying out is well
within the capabilities of those with 2 year associate diplomas.

I see little value in graduate development programmes or engineer in
training schemes. The graduate with whether they have a 1 year certificate
of 4 year B.Eng is a trainee technician. If the services of an engineer are
really required by a business enterprise, than the trainee with the B.Eng
will quickly rise from registered technician to registered engineer. If not
then both will remain at the level of registered technician. Further more if
a dedicated, interested and competent technician is able to complete a task
properly without need of the enabling competences of a B.Eng then such is
not necessary for the task, and an engineer is also not required for the
task.

Further more if 50% is the average pass mark, then a 4 year B.Eng is
equivalent to a 2 year associate degree, and in turn that is equivalent to a
1 year certificate, and so on, until passing through an academic programme
is not really of any value. All that matters is have an individual actually
interested in doing the job properly. Hence independent technical check
against a code of practice.

Currently no real filtering system. By setting a low rung, many people can
get over it, but not everybody. But everybody has the opportunity if they
have the interest. Few people will complain about restrictive work practice
and unfair trading. The vast majority of people in the industry will thus
become registered. The task of the engineers is to transform data into
information and push that information down the hierarchy so that design
decisions can be made rapidly and be implemented. No one wants paper, they
want the cars, houses, hospitals, and schools. By progress filtering,
lifting the rung higher and higher, can achieve a greater level of
confidence that those at the top: the engineers, have the highest level of
competence. If start at the top, then the self appointed experts, the
engineers become a joke. The engineering associate even the drafters with 30
years experience, start wanting to know how some idiot became an engineer.

I don't believe engineers are doing their job properly. To filter them out
and increase their status as the IEAust wants to do. Then should start by
filtering at the lowest level, and see if they are all competent to be
registered design technicians. When the next level is introduced most will
not have the day to day experience to move up to the next level. Not
interested if they have potential, want to know if they have actually used
higher level theories in practice on real buildings which have performed
satisfactorily and done so under the supervision of someone with greater
experience and expertise. Difficult to assess when results pumped out of
computer programs. But its not the numbers it's the qualitative
understanding that is of greater importance. Get the theory right first,
then get the numbers right.



Regards
Conrad Harrison
B.Tech (mfg & mech), MIIE, gradTIEAust
mailto:sch.tectonic(--nospam--at)bigpond.com
Adelaide
South Australia





******* ****** ******* ******** ******* ******* ******* ***
*   Read list FAQ at: http://www.seaint.org/list_FAQ.asp
* 
*   This email was sent to you via Structural Engineers 
*   Association of Southern California (SEAOSC) server. To 
*   subscribe (no fee) or UnSubscribe, please go to:
*
*   http://www.seaint.org/sealist1.asp
*
*   Questions to seaint-ad(--nospam--at)seaint.org. Remember, any email you 
*   send to the list is public domain and may be re-posted 
*   without your permission. Make sure you visit our web 
*   site at: http://www.seaint.org 
******* ****** ****** ****** ******* ****** ****** ********