Website Banner. John Monash: Engineering enterprise prior to World War 1.

[Main Index.] [Bridges Index.] [Unrealised Arch Bridge Index.] [People Index.] [Abbreviations.] [Units & Currency.] [Glossary.]

Unbuilt Monier Arch Projects, page 6.

Sewage Aqueduct, Geelong: Monier Arch Tender.

All the sketches below are in Monash's hand.

Two tall piers, set wide apart, are surmounted by the horizontal line of the sewer conduit. The profile of the arch ribs, which were to run parallel to the conduit, one either side of it, is roughly sketched in, rising above the level of the conduit.

'Back of envelope' sketch elevation of typical span.
Reinforced Concrete & Monier Pipe Construction Co. Collection, University of Melbourne Archives.

The sketch plan shows the two arch rings, each 20 inches wide, and set 5 feet 2 inches apart. They are joined along their length by five cross-beams; and near the supports by diaphragms.

Sketch plan (distorted scale) showing clear span of 144 feet (43.9m).
Reinforced Concrete & Monier Pipe Construction Co. Collection, University of Melbourne Archives.

Overview.

RC&MPC prepared a design and tender for the Sewage Aqueduct across the Barwon River in Geelong. Monash acted as chief design engineer and the project was close to his heart. Detailed computations were carried out by assistant engineers S. J. Lindsay and J. A. Laing. The client was the Geelong Water Works and Sewerage Trust whose Secretary was an engineer, J. S. Sharland. The Trust's Engineer-in-Chief was R. T. McKay. Monash had had dealings with McKay the previous year when he prepared an unsuccessful tender for Geelong's Ocean Outfall Sewer, and had concluded that McKay was antagonistic to RC&MPC.

The aqueduct was a major structure. Monash's design comprised 16 sets of twin arches, with a rise of 22'-9" (6.9m), resting on piers spaced 150 ft (45.7m) centre-to-centre. The conduit acted as a beam, supported on props in the region of the piers and slung from concrete hangers elsewhere.

Rough sketch for calculation purposes, showing the profile of the arch ribs, springing from the tops of the piers. The sewage conduit passes above the pier tops, at which points it is supported by stub columns. Within the span, it is supported by tension members hung from the arch rib.

The left hand side of this illustration is JM's sketch showing half an arch nominally divided into segments for calculation purposes. I have faintly mirrored the outline to the right to give an impression of the full span. A.H.

4. The cross-section of the duct is a U shape closed by a flat top slab. 5. Diagram showing how the duct was to be suspended from the arch ribs by a long U-shaped hanger. The arms of the U were to be 12 inches square and contain eight 5/8 inch reinforcing bars anchored in the arch ribs.

4. Sketch cross-section of the duct.
5. The duct (shaded) was to be suspended from the two arch ribs (shaded) by a U-shaped hanger.
Reinforced Concrete & Monier Pipe Construction Co. Collection, University of Melbourne Archives.

RC&MPC's tender was £13,725; but the contract was given to the NSW firm of Stone & Siddeley at £18,450 for reasons that were not clearly expressed. Monash was incensed, but there is no evidence that he made any effort to have the decision changed. Stone and Siddeley's structure, with its reinforced concrete trusses somewhat like a small-scale Forth Bridge, still stands though it is unused and is the subject of controversy over its preservation. It can be seen from the west end of Leather St in the Geelong suburb of Breakwater.

The structural system is reminiscent of the Forth Bridge in Scotland; but on a much smaller scale and using reinforced concrete members.

Siddeley & Stone's successful competing design with reinforced concrete trusses. Considère System.

Historic images of Siddeley & Stone's Aqueduct may be found in the University of Melbourne Archives Image Collection UMAIC. Search under Record ID for UMA/I/6624, 6625, 6467, and 6468. Any enquiries to UMA regarding these images should quote Location Numbers BWP/24126, 24127, 24128, and 24129 respectively. Further images of the aqueduct held by UMA have Location Numbers BWP/24130 to 24132, and NN/1021. Images relating to Siddeley & Stone's precast factory, including ovoid pipes and a small loco, have Location Numbers BWP/24026 to 24032 and NN/840 to 842.

Timeline.

no date: Specification. Tenderers are told that they need not adhere to the layout shown in McKay's "sketch design", but spans must be between 100 and 200 ft.

On the back of this document is a 'back of the envelope' sketch of the arch and conduit (see above).

1912/07/17: Age. Tenders are called for an aqueduct 2400 ft long in either steel or reinforced concrete.

1912/07/22: JM to McKay. He sees that tenders are called. Does McKay have a reinforced concrete design or do reinforced concrete tenderers do their own?

1912/07/23: McKay to JM. They do their own.

1912/08/06-12: Set of initial calculations dated 6th to 12th including details of the hangers which carry the conduit near mid-span. It ends with a sketch profile showing a half-span of 72 ft. The radius is 172' until 32 ft from springing, at which point the rise is 15'-5". The centre portion has a radius of 138 ft.

1912/08/12-13: Second set of calculations. Page 5, entitled "Final Design", shows the rise above "actual springing" to be 22'-9". Coordinates for the profile of the arch are found by the bending moment method. The axial stress in the crown is found to be 325 psi and that at the springing 300 psi. The required radius for the central portion is 124' and that for the haunches 158'. A calculation to check the stability of the arch and conduit against overturning by a lateral wind pressure of 30 lbs/sq ft shows the resultant of dead weight and side wind to fall within the middle third of the contact area at the top of the pier.

1912/08/14-17: Third set of calculations, including a check on the lateral bending strength of the arch at the crown.

1912/08/16: Set of calculations for quantities.

1912/08/17: Fourth set of calculations. The first page is a memorandum from JM re the Pile Substructure, saying "This has not received sufficient consideration …" and giving directions for further theoretical investigation. Then follow calculations as directed, by an assistant engineer.

1912/08/18: "Agenda" by JM. This includes: "1. Mr. Laing to [further?] investigate pile foundations".

1912/08/19: JM's memo of a conversation with Mr. Good of Dorman Long. They have decided not to tender, so Good gave JM his opinion that a concrete version would cost £15,000 and a steel one £20,000.

1912/08/20: Set of calculations entitled "Cutting down of Quantities …"

1912/08/21: Fifth set of stress calculations to check the reduced dimensions.

1912/08/22: JM to Lynch giving instructions for a reconnaissance of the site and meeting with McKay. "It will be well to get in touch with Mr McKay and pump him as much as possible; but every care must be taken to keep secret the type of design we propose."

1912/08/24: Lynch's report on his site inspection and meeting with McKay. Access to the site over cultivated fields, and soft shorelines is difficult and the nearest operational railway station is South Geelong, 3 miles away. The Show Grounds platform is closer, but the Railways will charge for special use. Trial shafts sunk by the Trust show black soil, then silt, yellow clay, and blue clay, and then layers of shells and silt. Rock is expected at 26 ft in the first shaft and 30 ft in the second. There will be a danger of flooding if materials are stored close to the works. McKay insists on Jarrah or Ironbark for piles, but would prefer mass concrete foundations down to rock. Any piles would have to be driven to rock. McKay continually emphasised his strictness re the gauging of materials for concrete. "His idea of good concrete I gather consists of using plenty of cement. He says he hopes the Coy gets the contract and is sorry the pipe tender was put in in such a way that he could not recommend it. Generally I think he would be very strict in small items but manageable on larger issues." Lynch describes McKay as "very friendly".

1912/08/27: Sixth set of calculations entitled "Final Stress Computations". These seem to be the final "Final Computations"!

1912/09/02: RC&MPC to Trust. Letter of tender together with a description and justification of the design [see below]. Price £13,725.

1912/10/01: Sharland to JM. The RC&MPC tender was unsuccessful.

1912/10/02: Geelong Advertiser. There were 8 tenders, comprising 12 designs. The Stone & Siddeley design was "adopted after exhaustive investigation". It has 13 spans of 176' and 1 of 136', 16 ft above navigation level. "The design is bold and artistic, and far from being a blemish in the locality, it will be a unique specimen of work, of particular interest to the engineering world."

1912/10/03: JM to "My dear Gummow, If you have not heard it from other sources, it will perhaps interest you to learn that the contract for the sewer Aqueduct at Geelong has gone to Stone & Siddeley, the price being £18,450/-/-. This result is, to say the least of it, ludicrous; seeing that our tender was for £13,725/-/-. Having had previous experience with Mr McKay, I took all sorts of good care that he should be allowed no loophole whatever to turn down our tender. The terms and conditions of his specification were adhered to most literally, and to deprive him of any excuse of saying, as he did before, that our tender was in any way irregular, the tender included a clause to the effect that, if in any part of our designs and specifications there appeared to be any departure from the express terms of the specification, it was to be understood that the official specification should prevail and be complied with. There was also a clause in our tender to the effect that, if the engineer required any portion of our work to be strengthened or amplified over and above what was shown in our drawings, we were to be bound to do this on terms to be fixed by the engineer himself. Apart from this, no less than three Melbourne University Engineers working independently, went carefully over every portion of the design, and the ultimate design as sent in expressed the most stringent view of any of the three upon every aspect of strength and stability of every part of the structure. There is no doubt in my mind that the design was in every way unimpeachable. As lower tenderers, we had the undoubted right of being asked to state for what price we would be prepared to carry out any modifications which the Engineer might have thought to be necessary to make our design acceptable; but we were not approached in the matter nor given any chance. In short, we have not had a square deal, nor have any of the other tenderers. One does not care to use strong language in a letter, but you can quite imagine what I think of it, especially in view of Mr McKay's having given me, that Saturday morning at the Stock Exchange, a definite and unequivocal assurance that if there was anything about our tender which he did not understand, or did not like, he would undoubtedly at once send for me to come down to Geelong to give me an opportunity of clearing up the difficulty."

1912/10/05?: Geelong Advertiser. Letter to the Editor from "Citizen" calling for the release of all tender prices.

1912/10/07: Geelong Advertiser. An article entitled "Sewer Aqueduct Tenders" argues that the figures should be released.

1912/10/07: Mr Renton [?] of Moorabool St, Geelong to Alex Lynch mentioning that the letter in the paper was his.

1912/10/07 and 15: JM to McKay, requesting return of the drawings.

1912/10/10: News of the Week. Article including a drawing of Stone's design.

1912/10/16: Geelong Advertiser. Similar article entitled: "Sewerage Aqueduct Tenders".

1912/10/16: Geelong Times. "Geelong Sewerage Scheme. Tenders for the Aqueduct. The tenderers for the construction of the sewerage aqueduct submitted designs on the specification prepared by the Engineer, and in this they were asked to furnish stress diagrams and calculations, also to show, in detail, the position, diameter and percentage of reinforcement … Some of the tenderers did not comply with the conditions required, nor were the calculations and drawings complete. The figures were checked, and further calculations made from the drawing submitted, and in some instances the results were not satisfactory, particularly in regard to the stability of the completed structure. Seeing that the structure is to stand for all time, stability must be the first consideration. The prices shown below do not truly indicate the relative values of the work. There were variations in regard to the depth to which the concrete foundation was taken before being placed on the piles, and the cost of these in some instances was based upon timber as against concrete. The variations in the designs necessary in some cases to render the structure stable would have led to a considerable increase in cost above that stated in the tender. For the purpose of arriving at a recommendation to the Trust, the whole of the tenders were brought into line as regards to the factor of safety against overturning due to lateral pressure, the depth of foundation, the class of piles, and the percentage of steel reinforcement."

Tender prices as published with the above article. (Prices in £.)

Poole & Steele, Balmain, Sydney Steel41,250
Cowley's Eureka Ironworks, Ballarat Steel34,574
Taylor Bros, Melbourne(a)RC26,500
 (b)RC20,650
L. Messy-Rhine, Geelong(a)RC21,000
 (b)RC14,765
Siddeley & Stone, Geelong(a)RC24,655
 (b)RC20,040
 (c) (1)RC, timber piles18,050
 (c) (2)RC, r.c. piles, timber handrail  18,450
 (c) (3)RC, r.c. piles, r.c. balustrade  18,950
Peter Rodger, Melbourne RC18,820
G. H. Dunlop, Melbourne and Sydney   RC17,978
Reinforced Concrete Company [RCMPC](a)RC, r.c. piles13,725
 (b)RC, timber piles13,125

[Taylor Bros (b) and Messy-Rhine (b) did not comply with the specification as regards spans.]

1912/10/22: McKay to JM returning the drawings.

Top.


Monash's Description and Explanation of his Design
to accompany the tender.

2nd. September, 1912.

AQUEDUCT ACROSS BARWON RIVER - GEELONG
Description of Design
in
REINFORCED CONCRETE

GENERAL:-

The Aqueduct proposed in this Tender is 2400 ft. long arranged in 16 equal spans, each 150 ft., centre to centre of piers.- It has thus been sought to secure perfect symmetry in appearance and arrangement, while, as will be seen from the "general elevation" sheet 1, all piers are located well clear of the river beds. Each span consists of a pair of massive arch ribs, braced transversely by transoms and webs, from which the conduit is suspended, in the central zone of the span, and rests, in the region over the pier supports.

WATERWAY:-

These arch ribs have a 'rise' of 22'-9", and the springing is kept at a uniform distance of 7'-9" below 'invert level'. This does not, however, involve any sensible encroachment upon the waterway, as the obstruction to the total waterway by the small portions of the arch ribs which lie below flood level, is less than one per cent of the total and is more than compensated for by the fewer piers.-

LOADS:-

The structure is designed to carry, with a factor of safety of 4 for steel and 8 for concrete (28 days old) the following loads:-

I.Self load of the whole structure
II.Water load of Conduit flowing full.
III.Crowd load of 460 lbs. per ft. run.
IV.Wind Load of 30 lbs. per sq. ft.

FOUNDATIONS:-

The tender is based upon the employment of pile foundations, and prices are given alternatively for Concrete and for Timber piles. In order to ensure that the piles shall offer a satisfactory resistance, quite irrespective of the assumed position of the rock line, it is proposed that the specification for the pile driving shall be as follows: "The piles shall be driven until the yield under a 5 ft. drop of a 2 ton monkey (or its equivalent) shall not exceed one inch in 4 blows". This resistance, worked out by the formula of Professor Brik (see City of Melbourne Building Regulations for piled foundations) gives a factor of safety of 4 under full load.

PILES:-

Tender A is based upon the employment of Concrete Piles. This Company has had considerable experience in the manufacture and driving of Concrete Piles (notably in the long Railway Viaduct over the Port Adelaide River). Up to 30 ft. long the piles to be 15" × 15". Above 30 ft. to be 18" × 18". For the very deep foundations it is proposed to make and drive the piles 40 ft. long, and then prolong them in place to the full length, and then drive further to the required resistance.

Tender B is for Ironbark or Jarrah Piles in one length. While this is cheaper, the permanent character and greater strength of the Concrete Piles is recommended for favorable consideration.

The piles being cut off a little above natural surface will project about 4 ft. into the mass concrete of the piers. In the lower 5 ft. of the piers, two grillages (Monier system) are formed, to bond the concrete into a monolithic mass, acting as a combined cap for the group of piles.

ARCH RIBS:-

The arches will be designed so that the line of resistance falls in the geometric centre of the arch ribs throughout their whole length, and therefore, as the distribution of loading is constant, and uniform, except a negligibly small irregularity in the crowd load, there will be no bending moment stresses in the arch ribs. Nevertheless, the longitudinal reinforcements in these ribs are introduced and arranged to take up a very considerable bending moment, also thermal and shrinkage stresses, thereby greatly increasing the factor of safety.

The stress diagram (sheet 2) shows that the total crown stress under full load, including crowd, is 362,500 lbs; the area at crown is 920 sq. inches; therefore the stress intensity at crown is 394 lbs. per sq. inch. Similarly the stress intensity at the haunches is 353 lbs. per sq. inch.

These stress intensities are very moderate. Concrete of the strength specified (1:2:2) will have an ultimate compressive strength considerably above 3000 lbs. per sq. inch. Thus without taking any account of the reinforcements, the factor of safety is at least 8. But, in addition to longitudinal bars, the design submitted also embodies helical reinforcements, according to Considère's well known method of 'hooped' compression members, which more than doubles the resistance of the concrete to compression.

ABUTMENTS & PIERS:-

The arch thrusts at all the piers, when the aqueduct is complete, will entirely balance each other, so that all piers intermediate between the end abutments will be stable. But to minimise all risks of unbalanced thrusts, during construction, every fourth pier is designed on one side as an Abutment also, so that the whole aqueduct really consists of four separate and independent sections, each of four spans.

STABILITY under WIND PRESSURE:-

The ratio of length to width here adopted may raise some question as to the lateral stability of the design submitted. This has, however, been carefully and amply provided for, as shown by the stability diagrams.

Assuming, first, that the arches merely rest upon the piers and are not anchored to them in any way, the factor of stability under a 30 lb. wind (when conduit is empty - which is the worst condition) is 2.8. But, in addition to this, the arch ribs are so tied back to the piers, that the whole weight of the pier and the whole 12 ft. base (measured transversely) comes into play, and thereby the factor of lateral stability is increased to at least 4.0.

The very ample transom arrangements provided ensure, also, lateral and longitudinal stiffness of the arch ribs. The greatest unbraced length is only 14 ft., so that the ratio of length of column to least lateral dimension nowhere exceeds 8½. Thus, no 'long column' action is anywhere possible, as this does not come into consideration unless such ratio exceeds 18.

SUSPENSION SYSTEM:-

For the central 90 ft. of each span, the conduit hangs from the arch ribs; but this suspension is by no means in the nature of a mere flexible hanger system rendering the conduit liable to swing in the wind. Upon the contrary the wind pressure upon each 15 ft. section of Conduit (for which each pair of suspension brackets is responsible) has been carefully worked out, and the bending effect upon the 'hangers' has been computed; the hangers themselves being designed to have a stiffness ample to give complete lateral rigidity.

Apart from this, the conduit itself is designed to act as a horizontal girder - so that it has a rigidity of its own, independent of the resistance of the 'hangers'.

EXPANSION JOINTS:-

It is thought prudent, to provide - in the conduit - an expansion joint over every pier. In a span of 150 ft. our experience has shown that the longitudinal movement, under extreme range of temperature, is not likely to exceed ¾". Indeed, under the equalizing effect of the almost constant temperature of the sewage, there is never likely to be any great range of temperature - except when the conduit is empty.

Provision has, however, been made, as shown in the detail, (sheet 2) for a movement of 2 inches; the continuity of the conduit is broken over every pier, a 2 inch gap is left, which is filled with plastic bitumen, and the joint is covered by wrought iron plates. The joint is designed so as to be easily taken asunder at any time and readjusted.

GANGWAY:-

A clear headroom of 6 ft. 6 inches is provided throughout the whole length of the aqueduct, and the gangway is securely railed for its whole length. The cover of the conduit is built in situ, except for moveable cover plates, two of which are provided in each span. The main arches have been computed to take, in addition to self loads, a crowd load of 140 lbs. per sq. ft. over the whole span, simultaneously with the sewer flowing to its full capacity.

CONCLUSION:-

While, as far as can be anticipated, every probable criticism of the design has been attempted to be met, it is of course impossible to foresee every aspect that may arise. The object of this tender is to propound a design which shall be in every respect in consonance with sound engineering principles, and it is to be taken as an essential feature of this tender, that this Company is prepared to embody in the design any further provisions in the direction of complete safety and efficiency that the Trust's Engineer may consider desirable to ensure a satisfactory and successful structure upon the general lines indicated in these papers.

[signed] John Monash
M. Inst. C. E.
2/9/12

Top.


Monash's "Summary of Stress Calculations".

The following document accompanied the tender and is in JM's handwriting. I have attempted to reproduce the general appearance of the original calculations, but have made the following changes:
Formulas have been expressed on a single line, requiring insertion of extra brackets, etc.
JM and colleagues were not in the habit of placing a zero before the decimal point in numbers less than unity. These zeros have been inserted. Thus JM's ".3" becomes "0.3".
JM often indicated 'square inches' and 'square feet' by drawing a small square followed by the inch or foot sign ( ' or " ). These units have been typed out in full or as 'sq ft' etc.

Sep.2/12

Aqueduct across Barwon River
Geelong.

Stress Calculations.

Total Load per Span (portion resting upon Arch Ribs.)  
Concrete dead Load270,000lbs.
Sewage in Conduit109,700 
Crowd  69,100 
Total  448,800lbs
Load of Half-Arch is therefore224,400lbs

The centre of Mass of this load is 35'-4" from the centre of span.
Whence, since Rise of Arch is 22'-9", by triangle of forces, we have:-

Horizontal Thrust at Crown= 362,500 lbs.
Resultant Thrust at Haunches= 425,000 lbs.

Area of Cross Section of both Arch Ribs, at Crown, is 920 sq. inches
Therefore Stress Intensity (maximum) is 394 lbs. per sq. inch.

Area of Cross Section at Haunches, (neglecting the connecting web) is 1200 sq. inches,
Therefore Stress Intensity (maximum) is 353 lbs. per sq. inch.


Girder Strength of Conduit
The Conduit is supported every 15 ft.

The Load per foot so carried is:-self load640lbs
 sewage730 
 crowd 460 
 Total1830lbs per ft. run

Bending Moment is
1830lb × 152 × 12/10 = 496,000 inch lbs.
Overall depth of Conduit is 57"
Effective girder depth is 50"
Tensile Reinforcement (six bars ½" dia) = 1.17 sq. inches
Therefore Stress in steel = 496,000 / (1.17 × 50) = 8400 lbs per sq inch.


Conduit as a lateral girder under Wind pressure of 30 lbs per sq inch

Load = 4'-9" × 15' × 30lb = 2135 lbs
Moment = 2135 × 15 × 12/10 = 38400 inch lbs
Width is 44"; Effective girder depth = 39"
Therefore Stress in 0.3 sq inch [of] steel = 38400 / (0.3 × 39) = 3300 lbs per sq inch

Suspension System.
All Hangers are designed the same as Central Hanger, which is the longest, and so the most heavily stressed.

Area exposed to wind, per Hanger,

due to Conduit15' span × 5' = 75.0 sq. ft.
due to hanger10' × 9" =  7.5 
  82.5 x 30lbs = 2475 lbs

Moment due to this wind load is 168,750 inch lbs at the point where the Hanger connects with Arch Rib.

To take this moment, there are provided two beams each 12" × 9" each having 6 bars ¾"dia.

This gives Stress Intensities in Steel well within the permissible limit of 16000 lbs per sq. inch


Vertical Load on Hangers is, as above, 1830 lbs × 15' = 27500 lbs

To carry this load from Arch Rib we have 12 bars ¾" = 5.28 sq inches

Therefore Stress in steel is 5200 lbs per sq inch.


Stability

(a) about plane of Springing

Consider each half span separately

Area subject to Wind pressure = 500 sq. ft.

Centre of action, above Springing, is 11'-7½"

Overturning Moment = 500 × 30 × 11'-7½"= 174500 ft lbs.
Vertical Weight (see above)= 144000 lbs

[The figure 144000 appears to be the self weight of half a span (without sewage or crowd load) plus half the weight of one pillar 6'-10" wide down to springing level, at the top of the pier.]

Therefore Arm required for Stability = 174500 / 144000 = 1'-2½"

Available width, at springing being 6'-10" the arm actually available is 3'-5"

Therefore Factor of Stability = 3'-5" / 1'-2½" = 2.83

In the above we neglect the fact that both arch ribs are securely anchored back into the pier mass, and so cannot hinge about the springing line as above assumed.

(b) about base of Piers.

Arm subject to Wind Pressure is, now, 19'-0"

Weight of half arch = 144,000 
of half pier =  50,750 
Total194,750 lbs

Overturning Moment

500 sq ft × 30 lb × 19' = 285,000 ft. lbs
Arm required for Stability = 285,000 / 194,750 = 1'-5½"
Available width of base = 12' and of arm = 6'
Therefore Factor of Stability 6'-0" / 1'-5½" = 4.


Figure of the Arch.

The Stress Diagram on Sheet 2 of Drawings [not located in RCMPC records] shews the stresses - both as to amount and directions - in each part of the Arch Rib. This diagram is based upon an approximate preliminary estimation of the weights of each portion of the arch; - and from these figures the centre of mass has been deduced. - Nevertheless, these calculations require still greater refinement, and a slight modification on the figure of the arch may be necessary, to achieve the result aimed at of securing that the line of resistance falls in the exact geometric centre of the rib.

The Arch rib is reinforced with 8 bars ½" dia on each side and thus the safe Bending Moment Strength of the Rib, at its narrowest portion (i.e. at the crown) is over 500,000 inch lbs. - There are no circumstances, however, which could bring about any material bending moment in the arch ribs themselves.

[signed] THE REINFORCED CONCRETE & MONIER PIPE
CONSTRUCTION CO. PROPY. LTD.
pp John Monash
M. Inst. C. E.