Are engineers becoming too conservative?

Are engineers becoming too conservative?

ARE ENGINEERS BECOMING TOO CONSERVATIVE?

It’s the most common question asked of every engineer when he or she gets to site: The moment the steel fixer believes the slab has too much reinforcement, or the steel rigger thinks all the steel beams are bigger than they’re supposed to be, the sarcastic question rings out, “What are we building here, mate?  The Sydney Harbour Bridge?”  Everyone thinks they’ve been lumped with a conservative engineer!

Stereotypes aren’t always appropriate or deserved, but they are hard-earned and established over time.  The general stereotype levelled against engineers is that we’re all a conservative bunch who over-design everything.   Being over-conservative with our designs or using overly-conservative factors of safety, the presumption is that engineers are risk-averse and more concerned with covering their backsides than designing efficient structures.

Is there any truth in this? Is the criticism warranted, or is there a bigger picture at play here?

It’s a tough tightrope for engineers to walk: You’re criticised if a structure is too deep, too wide, or too heavy – and yet you’re the first in the firing line to be questioned or sued if the floor ends up being too bouncy or if the structure displays cracks. Or, worst of all, you’re charged with criminal negligence if the structure fails and people are injured or killed.  Can you tell the difference between a conservative engineer, and one who’s more aware of the risks?

More interestingly, there’s a recent, growing perception amongst some in the architectural and building community that engineers are more conservative than they used to be.  It’s born out of a perceived observation that structures are bigger than they previously might have been in the past.  A great example of this is with concrete slabs:  Architects who were working with engineers in the 1980’s and 1990’s remember when slabs spanning four or five metres were typically just 160mm thick, and are now questioning or criticising engineers when the same slabs today are being designed at 200mm thick!

If there’s a perception that conservative engineers are now more conservative than they used to be, then what’s changed in the last two decades to trigger this?  Has there been a decline in engineers’ skills and abilities? Are the construction materials not as good as they used to be?  Have the design codes changed to make things more onerous?  Have the accepted factors of safety increased?   Have engineers lost certain knowledge or tricks?  The answers might surprise you…

First of all, before we address some of those questions, let’s come from the other angle and look at what’s improved for engineers in the last few years:

 

Concrete and reinforcement

Concrete being poured on site

Quite simply, concrete is better today than it was in decades past.   Higher strengths can be achieved, and better technology also exists in the form of admixtures and additives that improve the concrete’s workability, performance, and durability.   Strong advancements have also been made in the field of waterproofing technologies.  (Although be under no illusion – all the Penetron, Xypex, or Caltite in the world can’t and won’t make your concrete waterproof if it cracks excessively due to shrinkage or flexure).

Similarly, back in 2002, the yield strength of Australian-made reinforcing steel increased by 25% from 400MPa to 500MPa, allowing engineers to squeeze extra strength out of our concrete slabs, beams, and columns.

 

Steel and timber

A steel universal beam and a piece of structural pine

Australian-made structural steel is stronger and more efficient than it was in decades past.   In 1998, the yield strength of most hot-rolled products (i.e. Universal Beams, PFC’s, angles, etc) was increased from 250MPa up to 300MPa and 320MPa, allowing engineers to squeeze extra capacity out of their beams.

The stress grading of natural timber has also become more efficient and reliable since the late 1990’s. Visually graded radiata pine – the building block of most domestic construction in Australia – used to be assigned a stress grade of either F5, F7 or occasionally F8, whereas the new machine grading technology (MGP = Machine Graded Pine) allows pine to be more reliably graded, and for higher strength pines to be sourced for structural purposes.  (e.g. MGP12 or MGP15).

The last 15 years has also seen the cost of engineered wood products (e.g. LVL and LGL, and more recently CLT) come down tremendously, and so LVL joists and beams are now being used extensively throughout the industry.  LVL’s are now also coming in higher stress grades, allowing them to span further or support more load than was the case when they first appeared on the scene in the 1990’s.  Compared to natural pine joists, LVL’s are straighter, do not warp, and can have better performance characteristics for strength and deflection.

 

Analysis tools

Conservative engineer - modelling deflections with FEA
Finite Element Analysis software (FEA) being used to model deflections in a one-way banded slab. (Image courtesy of SkyCiv)

 

The computer and software tools available to the engineer today permit significantly more detailed, accurate, and advanced analyses than was ever possible in the past.   Even in just the last 15 years, the development and introduction of affordable Finite Element Analysis (FEA) software and non-linear analysis programs into the humble consultant’s office has allowed engineers to analyse concrete slabs or plate elements far better than the old 2D strip analyses they worked with up until the late 1990’s and early 2000’s.

This means we can push the envelope with our structures and get results that assure us the beams and slabs can span further or deflect less than our hand calculations would otherwise have told us in the past.

– – – – – – –

So, with the above as background, why is there a perception that engineers are over-designing?  Why is there a perception that we have more conservative engineers these days than might previously have been the case?

Well, when it comes to concrete, the reason is pretty simple:  The design codes we used in the past were……flawed!

Reinforced concrete has been around for over a century now, but it wasn’t until the mid-1990’s that engineers started to better notice and appreciate the long-term properties and behaviour of concrete structures.

To avoid a technical or verbose description, we’ll explain this in very simplistic terms: Suspended concrete slabs and beams effectively deflect twice.  First, they deflect the moment they’re loaded up.  (We call this the initial or short-term deflection).  This initial deflection is instantaneous.  Then, they continue to slowly deflect further over time, under the sustained loading.  We call this the long-term or creep deflection.   Creep deflection can continue for up to 30 years after the structure is poured, and its magnitude is usually greater than the initial deflection.

To put some simple numbers to this: A newly-built concrete floor slab for an office building might be fitted out with ceramic floor tiles on top; air-con ductwork and ceiling panels suspended underneath it; partition walls installed across it; and then get loaded up with furniture and a compactus unit.  The slab might deflect, say, 12mm under this newly applied loading.   Then, slowly over the next 30 years – even though the load on the slab has not changed or increased – it might deflect an additional 20mm.  The final, total deflection of the floor slab (in this hypothetical example) ends up being 32mm!  It’s the final 32mm deflection that the engineer has to design for and to future-proof the structure for in 30 years’ time.  Not the 12mm that occurs early on as the builder walks off the project on his last day on site and accuses the engineer of over-designing the concrete!

Deflected shape of a deformed slab ( Are engineers becoming too conservative? )

Engineers thus have to proportion their concrete slabs and beams to account for this long-term deflection.   There’s no point having a roof slab that holds its line for the first year or two, but after 20 years is drooping and sagging in the middle, collecting water and causing serious ponding and waterproofing problems!  Similarly, consider a cantilevering concrete balcony: Picture the top, trafficable surface being tiled on a screed with a gentle fall back towards the door threshold, where there’s then a slotted drain to take away the surface stormwater.  If the balcony is not correctly designed and proportioned for long-term deflection, the cantilever will start to droop at its outer edge over time, thus negating or even reversing the fall and preventing water from getting to the drain.  This is precisely the reason why so many apartment buildings built 10 years ago and older often require expensive remediation and waterproofing repairs – either addressing non-functioning falls, or repairing membranes and flashings that were damaged and torn due to long-term deflection.

By the 1990’s, engineers noticed that many of the flat plate slabs designed and built in the 70’s and 80’s were deflecting excessively, particularly in commercial office buildings.  Slabs were sagging and drooping significantly more than their calculations had predicted, causing all manner of problems for the building occupants.  (Occupants felt themselves walking downhill and uphill when walking along corridors in office buildings; furniture and desks in office fitouts would not sit level; wall tiles and finishes in the toilets cracked; flashings at slab and wall interfaces were torn; ponding occurred on roof slabs; cantilevered balcony slabs would no longer fall or drain in the right direction!)

The culprit here was the formulae for stiffness that engineers used to calculate and predict deflections. Engineers were designing correctly to the concrete code (AS3600 and its predecessors), but it turns out the formulas in the code itself were flawed and could under-estimate the long-term creep deflection in certain circumstances.  The first attempt to correct this in Australia was in 2001, when the stiffness/deflection formulas in the code were adjusted – resulting in slabs and beams becoming thicker or deeper than they might previously have been.   The formulas were tweaked again in 2009 as the industry responded further to ongoing research, testing, and field observations.  And, most recently, the concrete code got a significant re-write in 2018 that further adjusted the formulae for calculating stiffness and deflections.   The result?  Yes, all other things being equal, today’s slabs and beams will likely be thicker/deeper than an engineer would have specified them to be 10-30 years ago.

 

Graphic depicting how the formula for calculating stiffness has changed over the years ( Are engineers becoming too conservative )
The evolution of design: The above shows how the SIMPLIFIED formula to calculate a concrete’s beam stiffness has changed in subsequent editions of the Australian concrete design code (AS3600) as the industry responded to observations that predicted deflections were being under-estimated.  The 2018 edition of the code went a step further by replacing the long-established Branson’s equation for effective stiffness with the more onerous but seemingly more appropriate Bischoff’s equation.

 

But it’s not just the concrete’s thickness that’s been affected.  Similarly, today’s concrete code stipulates significantly more reinforcement in slabs than was previously the case.   Up until the early 1990’s, it was common for slabs to only feature top reinforcement in the negative bending regions over supports.  As such, large portions of a suspended slab would have reinforcement in the bottom layer only.   However, with the benefit of hindsight and now realising that this often lead to increased shrinkage strains and thus greater deflections, it’s now almost de rigueur to specify top reinforcement throughout the slab.  It’s not a case of engineers being more conservative or simply adding more steel to a slab so that we’ve got something to walk across when we come to inspect!  No, rather, today’s code has a better understanding of how to combat and resist concrete cracking and shrinkage issues, and it thus stipulates higher reinforcement rates than previous versions of the code used to require.  And we haven’t even yet mentioned the more stringent requirements for fire rating and durability that exist today!

What all this means is that – generally speaking – concrete slabs and beams being designed today will be thicker, deeper, and more heavily reinforced than what engineers would have designed them to be prior to 2001.  (And then subsequently again in 2009 and 2018 as the concrete code continued to become more conservative so that it dealt appropriately with long-term deflection).

The full benefit of this won’t be realised for another 10-20 years, when maintenance and remedial repairs on excessively deflecting slabs will hopefully start to become a thing of the past in Australia.  In the meantime, don’t give your engineer too much grief if your slab is an inch or two thicker than you used to see 20 years ago!

[BTW…the above discussion does not imply that all concrete floors built before 2001 will have deflection problems.  There are other parameters and variables that influence deflection (e.g. span, reinforcement, strength, and applied loads), and so not every floor will be susceptible.] 

The above discussion has focused on concrete, but similar developments and changes have occurred with other materials and structures. Civilisation’s construction history has been to build something.  And, if it falls down, we re-build it again, only this time a little bit stronger than we did the last time.  That is essentially what has happened with Australia’s design codes as the industry responds to structural failures, collapses, and research.   As some real examples, the code governing the design and construction of retaining walls now results in bigger footings and stronger walls than was the case thirty years ago – all because of older retaining walls either leaning or falling over in the last two decades.  Design wind loads have generally also increased recently in response to more data being collected and researched, and peak storm events occurring more often.  And strength and performance requirements for structures have become more stringent (i.e. conservative) in response to changes in the earthquake code, and for fire-resistance levels…..both as a result of disasters and tragedies that have occurred in recent decades.  No, it’s not the engineer, but the codes we work to are calling the shots.

– – – – – – –

Of course, there are other factors at play here, and several other good reasons why an engineer might design and size a member larger than you’re expecting.   A degree of robustness and flexibility is required to the design, and both engineers and their structures have to roll with the punches as things unfold on site.  After all, we frequently encounter design changes during the construction of projects, and these changes can compromise or impact the strength and performance of the structure.  Here are just a few solid examples of change requests that engineers regularly encounter on projects after the design is finished and construction is underway…

  • Floors that were originally shown on the architectural drawings as being carpeted or with lightweight timber flooring on top are changed to being tiled – suddenly adding at least an extra 100kg/m2 to the dead weight on the slab by the time you account for the screed and ceramic tiles. Similarly, roofs originally shown and specified as being lightweight, steel-sheeted get changed to being clad with terracotta tiles, thus significantly increasing the weight on the rafters and the weight transferred down onto the supporting walls.
  • Lightweight stud walls get changed to brick.
  • The client changes their mind for the bathroom and elects to add a huge, deep spa bath.  (It’s amazing how often this request comes through the morning after a big spa is featured on the previous night’s episode of The Block or Grand Designs or some other similar home renovation show!)  This can easily add up to a tonne in weight on the floor once you account for the water.  Even more if the spa features a heavy marble tile surround!
  • Supporting walls get moved by a few hundred millimetres to better accommodate joinery units, resulting in increases to spans.
  • Architects elect during the build to recess pelmets or lighting tracks into the slab’s soffit, thus reducing the slab’s strength.
  • Plumbers elect to chase or rout a pipe into the guts of the slab rather than suspend it under the slab.
  • The client elects to add a screed or topping to the slab to accommodate hydronic heating.
  • Ducts or risers not shown on the architectural plans have to be cut or penetrated through the floor structures, resulting in unforeseen voids and weaknesses.
  • The builder asks if the slab can be reduced by 15mm because he got his levels wrong.
  • The concrete gets poured on a hot day and the concrete delivery truck driver adds water to his mix on the road, thus weakening the concrete.
  • The architect requests a penetration through the beam so that services can pass through the beam rather than hang underneath it.
  • And….. well, you get the idea.

A good engineer would never endorse or encourage over-design, but a design should also be robust and flexible enough to cope with all of the above compromises and manipulations that occur during construction.   If the architect or builder wants to introduce a last-minute change on site, it’s a better outcome for everyone if the engineer can say “Yes, I think we can accommodate that,”  rather than, “No, we’ll have to re-design everything” or “You’ll have to demolish this and start again!”  And the client then gets hit with a variation for the additional design and building costs.  Not to mention the costs of delays to the build.  It’s not necessarily conservatism to design some latent robustness into the structure.  It’s experience and being cognisant of the realities of construction.

An old architectural plan from the 1940's
Prior to the rise of “open plan living” in the late 1990’s, architectural layouts in houses typically featured lots of smaller, isolated rooms surrounded by four walls – as illustrated in the above design from the 1960’s.  Such layouts provided ample support for the slabs and beams that landed on them, as well as keeping the spans down.  More modern and current architectural trends opt for larger, combined areas that challenge the structure.

 

Structures (and hence engineering design) are also responding to changes in architectural trends.   For example, the proclivity for “open plan” living means there are less supporting walls for suspended slabs and roofs to sit on than could once have been relied upon.  In short, the floors and roofs now need to span further than they used to!  The structure needs to be deeper/stronger to deal with this these days, although many in the architectural community have yet to update their expectations.  (We explored this in our previous article, “Getting the floor zone right – why the old rules of thumb no longer apply”).  The desire for higher ceilings in houses drives a need for floor zones to become thinner.  Houses now typically feature more glass to their facades, triggering more stringent deflection requirements to the transoms or floor beams that support them.  Concealed box gutters now cut into roof structures, reducing their effective depth and strength.   Off-form concrete is flavour of the month again, as well as polished concrete floors, thus putting pressure on engineers to make sure the concrete doesn’t display shrinkage cracks.  No one wants to see control joints or slab edges or have step-ups into bathrooms, etc, thus requiring more expensive detailing by the engineer.   All of these architectural developments and trends require structures to be more robust to accommodate them, and thus structural depths and sizes may seem to be bigger than what the industry typically accepted 20, 30, and 40 years ago.

Ultimately, engineers, architects and builders have to work together to achieve multiple common objectives:  An efficient structure; an aesthetic and functional outcome; and an economical construction.   Oh….and it has to stand up for the next 50 years without any defects, dramas, or disturbances.  ?

Cheers,
AD

PS – Agree with what you’ve read?  Had an experience where you disagree?  You can leave a comment in the Reply box at the bottom of this page.

The Working Engineer logo

Did you like this article?  Why not subscribe for more?

Subscribe for free and receive an email notification when we publish new articles...

Invalid email address
We promise not to spam you. And you can unsubscribe at any time.

 

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top