You are on page 1of 15

Integrating

Studies (16MMB501)
Computer Supported Design Report- Group L
Max Douch- B519767

Mark Fare- B523286

Josh Rice- B520877

Phil Yeung-B525185

Initial Stage
The very first design we did was a simple block like crane. It consisted only of 10 members,
a rectangular base and then a triangular arm (Fig. 1). The maximum PI this reached was
23.22.

When analysing the model, in the corner we removed two elements. There was no force
going through them so all they were doing was adding weight and not actually increasing
the ability of the crane to carry a greater force. We also changed the horizontal arm.
Rather than just being a triangle made is so there were alternate facing triangles aligned
next to each other (Fig. 2). This didnt have the same effect as we were hoping for as this PI
only reached 28.75.

Seeing as we didnt have the success we were hoping on the horizontal we kept it the same
as the first time and widened the base (Fig. 3). We still kept it simple for the time being to
get a general idea on how it all worked. The PI went up again to 33.88.

The next thing we did was to combine these two factors. We widened the base and added
in the more complex horizontal arm in (Fig. 4). Once the elements had been changed the
value of the PI went up hugely to 165.76.

Having seen the effect the triangular structure had on the arm of the crane, we decided to
try it with the vertical support. We kept it simple with only a few (Fig. 5) on both parts. This
again boosted the PI up to 247.55.

We decided to stick with this structure to improve from then on. We made the triangles on
the arm thinner (Fig. 6) as that way the force is distributed so you can carry a greater weight.
We didnt want to make them too small or we thought the mass would go up by too much
and therefore reduce the PI. This boosted the PI up to 294.


Fig. 1- PI=23.22 Fig. 2- PI=28.75

Fig. 3- PI=33.88 Fig. 4- PI=165.76


Fig. 5- PI=247.55 Fig. 6- PI=294


Individual Stages
Max Douch

From the previous work, we had a crane that had a PI of 294 and wasnt deflecting into the
shaded area.

The first thing I did was to see what affect widening the base had. As you can see from Fig.
7, it could hold a greater load but the weight was hugely increased (almost doubled),
therefore the PI was lower.

I moved the base back to its original width and then added more support to the vertical
column. Initially this gave me a PI of 276.39. The hollow circular elements and solid square
elements have very similar I values, therefore in buckling, they have a very smiler value.
However in tension, due to the area of the square elements, they can hold much more force
in tension. For all the bars that were failing in compression, I changed to a 12mm circular
element and for those failing in tension; I changed to a 5mm square element. After
changing all of the elements, see Fig. 9, the PI didnt go over 276.39.

For the next development I reduced the number of bars on the vertical column as they dont
carry any force. The original PI was 12.23. The maximum I made was 137.47, so the
reduction in bars didnt work. Adding in another element made the PI go down again, the X
shape doesnt seem to work.

I made it into a triangle shape and the PI made it up to 138.95 (Fig. 12). When I switched
the triangle the other way round the PI reached 182.66. When more elements were added
in the PI went up to 274.37.

I then concentrated on the horizontal. I made it the same style as the column and reached a
PI of 267.67. By switching the configuration and removing the end elements it made it to
237.7.

I started heading more towards the original design as these techniques werent working.
From Fig. 15s design I reduced the elements on the horizontal but only managed a PI of 259.
I added horizontal supports in to the vertical column and that boosted the PI up to 303.

On the horizontal I added vertical supports due to the success of them on the vertical.
However it did not have the same success as the highest PI I could make was 293. I decided
to make fewer triangles and again, it didnt succeed. The highest PI I managed was 259.

I went back to my design with the highest PI and made some changes, adding supports in,
but it meant the weight was too great and I only achieved a PI of 262.
The next thing I did was to change the horizontal part of the crane. I went with the original,
regular triangular design but added supports into all of them. The first PI I got was 51.8.
The maximum PI managed to get it up to was 239.

Rather than have the diagonal supports in alternate directions, I decided to make them all in
the same direction to see if this had any effect on the PI. The PI with them in tension was
292 and in compression 235. When in tension the 5mm elements were suitable to hold the
weight whereas in compression I had to use the 12mm hollow elements. The 5mm
elements are much more cost effective than the 12mm elements so therefore when in
tension a much greater PI would be achieved.

I concentrated on reducing the weight of the crane to reduce the PI now. I had less diagonal
elements and made sure that they were in tension rather than compression to make them
most cost effective. The maximum PI I reached was 282. Although it was lighter, it didnt
hold anyway near the force that it could before, hence the lower PI.

When I then changed the vertical support to the same structure and the PI went up to 301.

Fig. 7- PI=270.71
Fig. 8- PI=276.39


Fig. 9- PI=257.98 Fig. 10- PI=135.47

Fig. 11- PI=130.7 Fig. 12- PI=138.95


Fig. 13- PI=274.37 Fig. 14- PI=267.67

Fig. 15- PI=259.98 Fig. 16- PI=303



Philip Yeung

Fig. 17

Fig. 17 shows the initial crane design created by our


group which had a PI of 294. Its configured with
triangular structures. The element which is 45degrees
to the horizontal is experiencing large amount of
tension and is about to fail.


Fig. 18

Fig. 18 shows my redesign of the crane structure.
I reduced the width of the overall structure and
added more triangular shapes into the structure
hoping to strengthen the crane and also added
additional elements at the corner (previous
failing point) to reduce the tension. Unfortunately
the PI didnt improve but dropped to 140 and the
failing point is still at the corner.

Fig. 19

Fig. 19 shows a slight change at the corner of the
structure (previous failing point) which is similar
to fig. 17 and the PI went up to 295 which is only
tiny bit higher than fig. 17. The failing point is still
at the same place. The way to stop it from failing at
that point is to change the shape of that element
from a tube to a square bar as square bar is better
at tension than a tube.

Fig. 20
Fig. 20 shows significant improvement after
changing the 6mm tube at the corner to a 10mm
squared bar and the Pi has gone up to 345. But
now the failing point at the corner has gone. But
there are a series of failing points at the vertical
which are experiencing compressive loads.
Increasing the diameter of the elements at the
vertical might stop it failing from compression.
Fig. 21
In fig. 21 I have shortened the distance
between the loading point and the vertical
support of the crane hence decreases the
bending moment and the PI increases to
399.25 which is about 36% higher than our
initial crane design.

Josh Rice

The initial design created by the group (Figure 22) had a PI of 294 and used a
structure based on elements in the shape of triangles for maximum support and
stability. The crane failed in tension with a 5mm solid square bar. Elements under
tension perform best when they are square bars and those that are under compression
perform better when they are hollow tubes.

For the next design iteration (Figure 23) I recreated a similar crane structure but
with less intermediate vertical bars to reduce the weight. I also staggered the horizontal
width between support elements to combat the differences in tension applied through
the loading. This simulation has given a PI of 254.09 but where each element is close to
failure so no real progress can be made without changing the design. The vertical
components of the crane structure are the next to fail and so for the next iteration I will
add more supporting structure inbetween the horizontal beams.

For the initial stages of this design (Figure 24 & Figure 25) the load created failure
points again in the vertical axis so I have added extra nodes between the elements to
increase support. The changes made in this iteration have given a PI of 272.07.

(Figure 26) PI of 265.28. For this design iteration I started the crane again and made
it symmetrical as I wasnt having much luck adapting the original groups design to
achieve a higher PI

(Figure 27) PI of 302.53 achieved by using minimal material and slanting the
horizontal beam downwards over the 1m x 1m block. I left a much smaller gap between
the crane and the block for this design, which also helped to increase the PI.

(Figure 28) PI of 361.84 achieved by moving some of the nodes and changing some
of the elements shapes and sizes between bars and tubes depending on whether they
were failing in tension or compression. I also decreased the space between the crane
and the box again.

(Figure 29) had a PI of 402.33. This PI was achieved by increasing the number of
support elements in the horizontal and vertical beams. For this design iteration I also
widened the vertical base to increase support.

(Figure 30) PI of 797.74 by changing thickness of 6mm diameter steel bar to 12mm
diameter (top left) steel bar for the diagonal component in a lot of tension as shown.

(Figure 31) PI 806.94. This design iteration gave me the best individual PI score. The
total cost of this structure was 0.43 and the mass was 0.87kg with a load to failure of
305.1N. This PI was achieved without a counterweight system such as those used in
tower cranes, therefore with some additional time and adding a counterweight
structure Im confident that the PI of this design iteration could be further increased.


Figure 22 Figure 23 Figure 24
= PI 294 = PI 254 = PI 272


Figure 25 Figure 26
Figure 27
= PI 272
= PI 265.38 = PI 303

Figure 30
Figure 28 Figure 29 = PI 797
= PI 361 = PI 402



Figure 31
= 806.94

Mark Fare

This is the initial design (figure one) was designed by the group this had a pi of 294 and we
used triangulation throughout the structure. This provided the structure with the optimum
strength ad members to the left accounted for the tension, and to the right the compression.
The triangular elements give a strong local rigidity, which provides an even and effective
distribution of structural stresses. Figure 32

With compression acting on the underside of the structure and tension on the top outside
of the structure the failure point predictably was the top upright which mated the right
angle of the crane, this was a 5mm solid steel bar which performs better in tension than the
hollowed 6mm circular bar type. Looking at the bars used, many were 6mm however
debatably there were too many 12mm bars which reduce the PI considerably over the 6mm
alternative.

My initial improvement to the design was not to immediately tackle the area of failure and
instead improve the base structurally this would expectedly reduce the PI of the crane but
as the percentage to failure values of all of the vertical section were high I decided to re
design this area, this reduced the PI to 240 but made the design considerably stronger in the
vertical section. My hope with this change was to improve the design as a whole so when I
was to tackle the element in failure I would get a large jump in PI. Figure 33

For my next two iterations I decided to continue with the trend of improving the vertical
section further, I did this by increasing the number of elements via horizontal members
which will provide some extra support via tension to the structure the aim was to reduce
the buckling effect on the inside exterior members of the structure by halving the length
effectively doubling the load that they can yield. This again slightly reduced the PI to 198 as
a result of the extra mass and cost factor with no increase in load to failure. Figure 34

In figure 4 I then changed the failed element from a 5mm bar to a 10 mm bar, this increased
the PI by more than 80 and the design now had a PI of 286 this revealed failures in the top
horizontal section of the structure, I then decided to look at what size bar and type I had
used in each element, to see whether any of the 12 mm elements could be reduced to 6mm.
Figure 35

From looking at the table I found that one of the tope members were at 12 mm, I reduced
this to 6mm. This increased the PI to 307. The reason the PI increased is because the 12mm
bar weighs 0.491 kg per metre and costs 0.246 per metre, whereas the 6mm bar costs 0.062
per metre and only weighs 0.491 per metre. As this element wasnt contributing much to
the load to failure downsizing this made a positive effect on my design. As a result the
failure still occurred in the same locations but with the design being more cost effective and
lighter it proved to be more efficient. Figure 36

I then set about making the same change to the horizontal section as I did to the vertical
section, this allowed me to reduce the bar size for the top elements as more distributed the
force of the load this combination increased my PI to 368. This increase was less than
expected. I then started to consider reducing the size and length of elements where possible.
Figure 37

Started to reduce the length of members where possible

Started to look at any members which had a low percentage of contribution to failure, saw
that 33 had 8% and was at 12mm so I reduced it to 6mm which increased the performance
index to 474 Figure 38

I then went about improving the design further, I achieved this but moving the nodes and
reducing the amount of material used in each element, each time this slowly improved my
PI I also found that the closer the structure is to the load, the smaller and more compact the
structure can be, this is due to the bending moment being far less the closer the element is
to the load point. Also a systematic failing occurred on the vertical part of my design, which
is a more encouraging sign that my crane was getting less over designed and more efficient.

Moved the node at the end of the design which reduced the overall weight marginally

Moved the structure closer to the 1x1m no go area to maximise the reduction in material.
Failure at more points PI of 518 Figure 39

A re design id possibly needed as this design is closer to reaching its potential with the
current material choices.
Possibly changing some thicknesses could increase the PI more

Research on crane designs will show areas for improvement and ways to maximise load to
weight ratio which will also attribute to reducing cost.

Figure 32 Figure 33 Figure 34


= PI 294 = PI 240 = PI 198

Figure 35
= PI 286 Figure 36
Figure 37
= PI 307
= PI 368

Figure 38 Figure 39
= PI 474 = PI 518
Crane final iteration
(Disclaimer) We unfortunately lost the 806 PI crane file, and whilst this was ultimately
the highest PI we achieved as a group we feel that the design as a concept had a lot
more potential and we ran out of time to improve the design further.

What we know and have learned

The Pi formula is:


Load to failure / Mass x Cost
(This means that mass and cost, because they can arguably be linked are
more important than the load to failure)
Structures need to be allowed to deflect slightly, the no go area must not be
entered, so some of the individual designs which are sitting on the no go area
must be changed.
The cost of the square bar is 30p per kg and the hollow tube is 50p per kg. So
these need to be considered in in optimising the element option.
Quoting the table the 6mm hollow bar has the lowest cost x mass factor at
0.008 while the 5mm square bar has a mass x cost factor of 0.011. this
means that any bar that isnt structurally dependent or not close to failure
must be kept at 6mm to achieve the highest PI
Designs will fail before the strongest bar types yield or buckle meaning that
the lower the bar thickness (where possible) the higher the PI
Tower crane research influence

From looking at tower crane examples we found that they were very efficient
with the load they could carry compared to the amount of material used, their
shape encourages the tension to be distributed evenly around the outside of
the structure whilst the central vertical and horizontal sections are more
structural rather than weight bearing. This is a feature that we found we could
utilise throughout the final design stages to increase the PI.
Normally tower cranes have a section left of the vertical section as a
counterweight, to mimic that we attached a third node in a fixed position with
no degrees of freedom to the base, this gave the horizontal strut of the design
a lot more strength. The only negative with this design is that the added struts
and lengths add more weight and these can be the ones which fail first,
increasing the thickness of the bar could solve the problem, however; reduce
the PI due to mass x cost factor.

In figure 40 we started with a solid design of a tower crane structure, All the
elements were at either 6mm if in compression or non-structural and of 5mm if
in tension and structural. This had a PI of 425
The Pi was still relatively low considering the design because the base was
too thick and due to the amount of elements in the design this increased the
weight and cost far beyond the load to failure advantage.

Ultimately this is the design we came up with and submitted and it was
achieved by reducing the amount of material used in the elements, while
keeping them all very small in size Shown in Figure 41. The failure points on
the design are in tension and this is possibly because the base has got so thin
that the support elements take too much of the effective load. Increasing the
material size of these from the 5mm to 10 mm square bar in tension reduced
the PI as a result of the length of the elements.
We as a group believe that this design could have taken us close to the 1000
mark for the PI if we found a way to optimise this design further, maybe by
changing a few thicknesses here and there. The best PI we submitted with the
design was 702

Figure 42 shows that the design doesnt deflect into the shaded area.


Figure 40 PI = 425
Figure 41 PI = 702

Figure 42
Appendix

QUESTION 1

QUESTION 2

You might also like