State of Play: 2017/wk40

The daily challenge came to an end on day 67. A daily challenge based on researching and reading would have greater staying power, but it doesn’t produce anything visual to communicate activity. All the production occurs inside the human mind, and makes the mind more knowledgeable. Any case back to random and irregular “state of play” posts.

Floor Loads

A few weeks back I was searching the net for weights of various household appliances, so that could use to design floor framing for Tiny House. Found some interesting sites by removalists, house movers, and storage facilities, along with moving guidelines for military personnel and government personnel. After a while thought, why am I doing this, we have a loading code? So  put that aside and went back to checking out dimensions of various appliances and drawing various layouts, attempting to collapse a caravan with bathroom facilities down to the smallest length. The starting point was an assumption that the most common size caravan with such facilities is 16 ft [4.875m], and width of caravans appears to be 7′ 9″ [2.360 m], and measurements using imperial units appear to be common for the vans. Based on a combined shower/toilet module 670 mm x 1025 mm, sink 370 mm x 370 mm, and cook top 480 mm x 370 mm, the minimum length appears to be 7 ft [2.145 m] . Shrinking the width of the bed from 900 mm down to 625 mm, can still squeeze everything into a 2.145m  long van, with width reduced to 1.390 m. The reduced width taking into consideration the width between wheel arches of some of the commercial  work vans.

Car License Weight Restrictions

Another issue was checking was the maximum weight of trailer which can haul on a drivers license. There doesn’t  seem to be any clear statement. The car drivers license is limited to 4.5 tonne gross vehicle mass (GVM), the gross combination mass (GCM) is dependent on the 1961 Transport Act. The act however only indicates that regulator can set limits, it doesn’t say what the limits are. Other documentation seems to indicate that the limit is the GCM of the actual vehicle, and given that the trailer is also considered to be a vehicle, for the trailer to remain a light vehicle, it needs to be less than 4.5 tonne. Therefore would appear that the maximum GCM is 9 tonne, though it would be preferable that hauled weight is less than the prime mover. Another limit is the 50 mm tow ball, which has a upper limit of 3.5 tonne, most do not appear to be installed for this load. Therefore the upper GCM limit drops to 8 tonne. Still another limit, is 750 kg for unbraked trailer, dropping limit down to 5.25 tonne. These are legal limits which need to be reconciled against the actual capabilities of the vehicle: for example a vehicle may require a braked trailer at less than the 750 kg limit, and a tow bar may be torn from the vehicle when attempting to pull less than 3.5 tonnes. So important to check the vehicle specifications.

A few years back when looking around at available at 1 tonne commercial vehicles, I hit the problem that the reference to 1 tonne was nominal, and the actual cargo capacities were only around 800 kg to 900 kg. When looking at the towing capacities of other vehicles such as utes, the towing capacity was typically less than 1 tonne. None of which was very useful, as the estimated weight of all my books, magazines and written work is 1 tonne to 1.2 tonnes. Such capacities also seem highly limiting to the building of caravans and camper vans.

Anycase looking more recently it seems a vehicle with a GVM of 4.5 tonnes typically weighs between 1.8 tonne and 2.2 tonne. Leaving around 2.3 tonne to 2.7 tonne for cargo. Also seems that the new generation of 4WD’s and utilities have a towing capacity of 3.5 tonne, though as this article shows, it limits the use of the vehicle. Seems typically 10% of the towed weight is placed on the tow ball of the vehicle, and this contributes to the vehicles GVM. So some playing around distributing the weight between a vehicle and a trailer, with typical GCM’s around 6 tonne to 7 tonne. This is important because Australian caravans are apparently amongst the heaviest is the world, due to heavy construction to handle unsealed roads and off-road touring, and there are few vehicles available which can tow them. Which leaves wondering as to whether the weight of a Tiny House is within the towing capabilities of available vehicles, given that tiny house construction is using relatively heavy construction materials compared to those used for caravans. Further checking also indicated that there are no structural design requirements for caravans in the Australian Design Rules (ADR) for vehicles: so chances are in the main no structural design of caravans taking place.

Building Code Limitations

In the meantime whilst collecting data for designing Tiny Houses and mobile offices, I received feedback on a council feedback on project. This led me to reading upon, legislation covering design of physical systems, as my contention was my project was and is not within the scope of the Building Code of Australia (BCA) whilst it may be within the scope of the South Australian Development Act and Regulations, the South Australian Building Rules have not yet been extended beyond a mere reference to the BCA. I have mentioned in earlier articles that BCA Class 10 is not a catch all for any structure which cannot be otherwise classified, and that BCA volume 2, a prescriptive guide for housing is inappropriate for the massive structures which cannot otherwise be classified as a building. Buildings have walls and a roof, such is not required for structures: it is the reason I proposed a national structural code and went further to propose a national technology code: and set up a specific purpose blog to comment further.

Anycase this also led me to take a look at the loading code AS1170, and look at where the loads came from, this in turn led me so researching human factors data, and checking some loads from first principles. My conclusion was that the loads in the codes are pulled out of a hat and bare little relevance to human factors and real world situations. This is not just an Australian thing, a few years back engineers in  the UK were questioning the loads mandated for office spaces. The basic problem is that the uniformly distributed area loads are typically too high based on unrealistic loading of a floor. The other issue is that we have moved to limit state probabilistic design, and still messing with safety factors. Under probabilistic design we do not need to keep the ultimate strength design action less than the breaking load, the load can equal the breaking load. Breakage will only occur if the actual action equals the actual breaking load of the component. The actual load varies upto the ultimate strength design action, with a low probability that it may exceed. The actual strength varies with a low probability  that the strength may be less than we expect. Our concern is that there is a low probability that the actual loading will exceed the actual strength of the component. Not some rubbish that the things is twice as strong as it needs to be: because the reality is we do not know how strong it needs to be when the load can vary, and we don’t know how strong it actually is because manufacturing processes do not produce uniform product all the time.

So it is time we got rid of the partial load factor of 1.5. That factor exists because of the soft conversion of our codes back in 1985 from permissible stress design to limit state design. It calibrates the results of the two codes to give similar answers. However the point of changing to limit state was so that could eventually get different answers, more economical and optimal answers. But its seems the loads are not based on any real research. For example when first changed we had a 0.8 reduction factor and 1.25 magnification factor for dead load: supposedly 0.8 allows for 5th percentile weight and 1.25 allows for 95th percentile weight, and so can adopt the appropriate value to check stability of a system. For example a plank with an overhang balanced by weight alone is not stable, according to the code, because the weight of the back span is only 0.8(SWT+DL) whilst the overhang is 1.25(SWT+DL). These values were seemingly pulled out off a hat, and then modified to the current values of 0.9 and 1.2, once again seemingly pulled out off a hat. In Australia we have no proper learned society, no proper journal of structural engineering, in which these changes are properly discussed by practising designers. {I Don’t wish to be disrespectful but Engineers Australia is little more than a modern day Rum Corp. It is far from being a proper learned society.}

The loads of concern to me are those for floors and those for barriers. I have previously discussed issues with the BCA and loading code problems with barriers: not the least the proper definition of a barrier. A barrier is not a fall prevention device, it is a traffic flow control device. A barrier is not something merely to be installed at the free edge of a floor from which a fall can occur. Looking at human  factors data, and attempting to assess loads on various clusters of humans, it seems clear that the floor loads are nonsense. For certain floor loads contain allowance for things other than assembly of people. However, loading from assembly of people should be separated from loading due to furnishing and equipment. The loads should be broken down into component parts so that design can better assess the load in a given situation. The description of situations in given in AS1170 are not overly clear, and importantly are largely concerned with buildings (things with walls and roof).

Having made this assessment about floors went back and made a check for a caravan with a floor area same as a 6m shipping container (Area: 14.4 m^2). For 1.5kPa, get total load of 21.6 kN [equivalent mass 2.2 tonne]. This is live load so partial load factor of 1.5 applies, this pushes it up to 3.3 tonne. Whilst the numbers maybe useful for structural design of the floor, it seems unrealistic that such floor would ever experience such load. More importantly there is an additional magnification factor for small areas, to ensure that the total load on a component part is not less than 1.8kN. It seems clear that a review of the loads is required to make them more realistic so as to produce more economical structures. Why are people choosing tiny houses? Whilst one reason is maintaining mobility, the other is that houses anchored to the ground are too expensive. If we adopted more realistic loads, then maybe the houses would be more economical.

Similarly, I also determined that barrier loads are also unrealistic. As far as I know 0.6kN is more than representative of the 95th percentile force that a person can exert horizontally, so to multiply it by a partial load factor of 1.5 is not acceptable. The industrial platforms code does not magnify the loads by 1.5 for testing, and the tests permit residual deflection: the result is that testing typically passes systems which calculations would reject.

There is another issue to consider, and that is in the main there are no mandated deflection limits in Australian codes. Deflection limits are a serviceability requirement, and in many situations, design for the extreme ultimate strength loads produces adequate performance at the lower serviceability or operational loads. For a structure like an overhead travelling crane, if deflections are too high, then components will jamb and the crane will not move: so deflection limits are imposed.

The lack of deflection limits results in the following situation. Client turns up at office and desires calculations for a floor beam, and  declares that they don’t want a steel beam because they bounce too much. We then have to explain that the material isn’t the issue, that the issue is whether or not floor vibration is considered in design, and further more it is the Australian Steel Institute which publishes guidelines for floor vibration not the timber development association. Therefore if steel beam is designed properly then it won’t have too much bounce. The timber framing code AS1684 span tables have made some token allowance for floor vibration dues to foot step impact. So floor framing taken from the code not likely to bounce too much. But if need calculations for a floor beam, then it is likely that outside the scope of AS1684 span tables, and therefore likely beyond the limits of timber. If it is within the capabilities of timber, then manufacturers tables for glulams and LVL’s can be used just like AS1684 span tables, no calculations required. If none of this suits then would need to design a floor truss, or to keep as a simple beam then change material. Steel is both stronger and stiffer than steel: so for long spans it is likely to be lighter. Which is the problem: if only design for strength then the beam will be strong but flexible. It also needs to be designed for deflection and vibration.

Platforms, Barriers and Crowds

Therefore if we reduce the mandated design loads, we also need to give more importance to deflections. More importantly we need serviceability loads not just ultimate strength loads. In checking for human factors found some useful websites and useful articles on crowds and history:

  1. Progressive crowd collapse
  2. Crowd crazing
  3. Static crowd density (general)

The crowd density article was interesting as I had previously looked in New Metric Handbook Planning and Design Data (Tutt & Adler 1997), and obtained maximum density of 6 persons per square metre, though I questioned the space allowed for each person: they stack people in a 2m x 2m square to arrive at their value. Seems packing in crowds can be higher than 6 persons/m^2. The above articles also gave some insight into origin of the 3kN/m barrier load for crowds, multiplied by 1.5 get 4.5kN/m. This value seems valid, on the other hand no barrier should ever experience such load: I reiterate the function of a barrier is to control, traffic it is not  a fall prevention device: design loads can always be exceeded. Well not only can design loads always be exceeded but the mandated minimum heights for barriers are questionable: these also do not appear to be based on anthropometric data.

Just to be clear vertical rails do not prevent climbing, they merely hinder. I have a scar above my right brow because around the age of 4, I climbed over the vertical rail child safety gate. Admittedly I built a staircase from my dad’s engineering text books, and rocks from his geology field trip: and I failed to plan ahead as how to get down on the other side. Fortunately it was at ground level and was meant to keep me in the house. Point is it is naive to believe that vertical rails make a barrier safe. It is also naive to believe a barrier of 865 mm or 1000 mm is of adequate height. A barrier should be solid, and around 1.8m above standing surface. Assuming maximum table/bench height of 900 mm, and minimum height of steps to reach light bulbs/globes also 900 mm, then minimum barrier height from floor should be 2700 mm. Thats higher than the typical ceiling height of 2400 mm. It is extremely naive to consider that code compliance is safe. Codes are written by committees,  the members of which have vested interests, interests which typically have little to do with public welfare. Remember houses are too expensive: the only loss of amenity we have to worry about is not getting in the first place. {NB: In England in primary school we were required to learn how to climb a vertical rope to the ceiling (anything from 3m to 6m high). Climbing trunks of trees and vertical rail park fences, and stone walls is a lot easier. As I recollect most of us weren’t any good at climbing the rope, and one of the two kids who made it to the top wasn’t able to get down. I also recollect the primary issue with park fences was kids getting stuck in the fence, especially their heads: not so sure that 125mm sphere ensures this won’t occur, and it certainly doesn’t stop things from being dropped from a great height, nor does a lack of a kick plate.}

Now clearly a barrier 2.7m high is going to block views unless it is transparent or fitted with windows. Additionally a solid barrier is also going to hinder activity at the location of the barrier. The BCA does not require a barrier where it would interfere with proper use of the space adjacent to. It mentions a stage. But really, a barrier is a problem for a stage? A glass wall can be fitted to the stage: reflection may be a problem, but appropriate lighting may be a problem. A barrier hinders operation of a wharf and loading dock: as cannot get goods on a off a ship. Really? Container cranes operate at heights greater than any barrier, and people get on and off ships via gang planks/ways. Though could be a problem for small boats.

A fishing platform on the other hand requires access to the water. Recreational fishing boats for example, have very low barriers at the working end of the boat, from photos around the internet, these barriers appear to be less than knee height. Other boats have no back wall, the water laps onto the deck, these boats provide easier access to the water for scuba divers. For game fishing seats with seat belts are provided, to keep person on boat. Design for purpose not to codes. So a land based structure, used as a fishing platform, shouldn’t have barrier as such would interfere with operation. After all most people fish from rocks, or banks of rivers, falling into the water isn’t considered an issue. Though if the water is tidal, then there is a risk of falling to the ground rather than into the water: but not likely to be there when the tide is out and cannot fish. So in such situation do we need a barrier and what purpose that barrier?

I reiterate that a barrier compliant with a code of practice does not prevent falling it just hinders the possibility. So some wharfs for example simply paint the edge of the dock yellow to highlight its location. Tactile indicators could also be installed to aid the blind in keeping clear of the waters edge. a raised kerb can also indicate the edge and also hinder wheel chairs rolling over the edge.  Where should a barrier be located and when  should it operate?

For a wharf, dock or fishing platform, a barrier can be set back from the edge, and can be provided with a gate, so that can access the free edge and the water when the tide is in. But have to intentionally open the gate to access such free edge. When people have access to the free edge of a platform they tend to sit down on the edge and allow their legs to swing freely below. If they have to stand, then some leaning rail or post may be advantageous to reduce fatigue. But under no circumstances should a barrier be installed which hinders the process of fishing or any other operation carried out from behind a barrier. So designing an appropriate device to hinder fall from a working platform is not a short term or simple matter. To be clear vertical rails do not prevent climbing, and horizontal members do not necessarily enable climbing. For example thin tensioned wires do not provide a stable climbing rung: though would be easiest to climb adjacent to a vertical support rail. But solid infill, vertical or horizontal rail infill, it all gets in the way of the practical operation as a working platform. Turning things on edge to operate through the rails is cumbersome, operating over the top of a high barrier is also cumbersome. Minimum space requirements is typically a circle 600 mm diameter, to enable a person to pass through a space. But point of a barrier is to stop people getting through. So we have two conflicting design requirements: stop passage and allow passage. The typical solution is a gate.

Any case whilst reviewing requirements for barriers, considering loading requirements the following made the news.

Soccer fans injured after stadium barrier collapses at Amiens

Which was interesting as had also recently read the following:

  1. Victoria Hall stampede
  2. 1902 Ibrox disaster
  3. 1971 Ibrox disaster
  4. Hillsborough disaster
  5. The Who concert disaster

There were some interesting articles about such disasters, conspiracy and controversy.

  1. Hillsborough: The Crush Barrier–A Smoking Gun?
  2. Viewer’s Guide to “Auschwitz – The Surprising Hidden Truth”

The latter one an alternate and unusual reason to look into crowd loads and crushing: amazing the stuff that turns up when searching for information.

Anycase the purpose of a barrier is to control the flow of traffic, there should be no reason to increase design load for barriers, if barriers are used to properly control the flow of people. Barriers however need to allow people to get into an area in a controlled manner, and then get out off the area in an equally controlled manner. The barriers fail, if they permit entry in a controlled manner, but hinder escape in a controlled manner. Furthermore barriers are completely useless if not properly maintained. Increasing design load is not the proper way to allow for poor maintenance.

Rainwater Barrels

After looking at barriers and loads, went back to looking at information for tiny houses. Including cost of materials from hardware stores. Also took another look at minimum land requirements for agriculture, UK allotments, and Russian Dacha. Then issues of rainwater barrels: seems in the USA such barrels are considered illegal in some states; seems crazy.

Is rainwater harvesting illegal?

Though I am in South Australia the driest state on the driest continent on earth: so water is important.  Though this is a bit of an exaggeration:

The biggest rain barrel you have ever seen

Not a typical collection tank in Adelaide. Adelaide has tap water, and before about the 1970’s rainwater tanks were popular, as the water was considered better than Murray River water. Back then lots of adverts for zeolite water softeners and all kinds of filters for tap water. Then water filtration plant built, and a few problems in mining towns with lead contamination of rainwater led to reduced use of water tanks. Rainwater is now classified as grey water, not recommended for drinking. Though some people still reckon rainwater tastes better than tap water: got rid of mud and replaced with chlorine. Anycase a large rainwater tank in suburbs likely 2000 L not 20,000 L. The larger tank more likely found in rural areas where tap water is not available and neither are sewer mains, and also have to use septic tanks.

Also rather than rain water tanks being illegal, highly likely to be required to install a rainwater tank and not to collect water, but to prevent flooding. Yes! in driest state on driest continent flooding is a potential problem. We oscillate between drought and flood. We refer to detention and retention tanks. I can never remember which. Detention I think is temporary and retention permanent. Some councils provide orifice plates to apply to the detention tanks, to control rate water is released from the tank. If install a carport or other canopy, likely to be required to install a detention tank .  The reason is that the stormwater mains were designed on an assumption of water typically being able to soak into the ground, and this was based on a certain percentage of the land being landscaped. With gardens being replaced by paving and larger buildings, less water is soaking into the ground and there is more surface flow, leading to the potential flooding as stormwater pipes not large enough: and nor are the natural creeks and rivers that the storm water may be directed to. Though some amy flow direct to the ocean. The detention tank detains the water on site, until a storm is over, and then slowly releases the water to  the main drainage system. The tank therefore is required to be empty before a storm, it is not for the purpose of rainwater harvesting: additional tanks are required for such purpose.


So a lot of things covered:

  1. Tiny Houses and caravans
  2. Trailer Sizes
  3. Car license weight restrictions
  4. Platform Design
  5. Barrier Design
  6. Design Loads
  7. Materials Costs
  8. Rainwater collection
  9. Farming, minimum land requirements

And more recent thing looked at was clearances between objects of differing heights, and the problems of creating a simple design rule and then all the inconveniences which result, giving rise for simple rule to be expanded into many rules.

Future Articles

So will post additional articles in next few weeks on:

  1. Site Visit to Price Boat Ramp
  2. Platform Design
  3. Review of Design Loads
  4. Clearances for space and avoiding objects falling onto one another.



  1. [07/10/2017] : Original