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

[JM Intro.]   [Main JM Index.]   [Bridges Index.]   [People Index.]
[Abbreviations.] [Units & Currency.] [Glossary.]

Monier Arch Culvert
Scott's Creek, Springfield

History

In April 1901, James S. Sharland AMICE, Engineer for the Shire of Springfield, was confronted with the task of building a road embankment across the gorge of Scott's Creek in the Shire. A large culvert would be required to allow the creek to pass through the embankment, and Sharland wrote to ask Monash & Anderson whether Monier concrete could compete with brick or stone for its construction. Responding to their questions, he sent engineering drawings of the site, about 7.5 miles (13 km) from Romsey Railway Station. He promised a sample of local sand, available at about 12/- per cubic yard, but doubted its suitability. He noted that the Council would insist on tenders, so M&A should be prepared to build the culvert as subcontractors.

M&A's engineering assistant, J S Gregory, prepared a drawing showing the arch of the culvert with a clear span of 12'-0" (3.66 m) and a rise of 5'-0" (1.52 m). Its thickness at the crown was only 2.5" (64 mm) and at the springings 5" (127 mm). Reinforcement is shown on the inside face (soffit) only. The strip footings sit on rock. The length of the culvert at this stage was 80 feet (24.4 m). Gregory also prepared quantities and a cost estimate of £146-5-0.

Cross-section as described in caption.

The left hand side of J S Gregory's drawing shows half of the cross-section of the culvert.
The right hand side shows half the entrance, flanked by rubble walls.

The difficulty of supervising work from M&A's office is indicated by Sharland's advice that he was intending to visit Scott's Creek on 27th, and if one of their staff left the city by the first train for Romsey, JBS could take him to the site and return him to the railway station to catch the evening train back.

On 30th Sharland wired that he had reduced the length of the culvert to 60 feet (18.3 m). M&A replied with figures of £160 for the whole job, or £85 for the Monier work alone, the royalty for the 15 cubic yards (11.5 m³) of mortar required being £12-10-0. Sharland then sent the contract drawings to the Public Works Department (PWD) for approval. After that, they would be on display to other contractors in the offices of the Municipal Association. He thanked M&A for their "kind and valuable assistance".

There was some delay, and after urging from M&A, Sharland went to see Carlo Catani, Chief Engineer of the PWD to learn that he had approved the plan, and it was sitting in the intray of William Davidson, Inspector General. Approval of contract documents was received on 29th May. Tenders were returnable by 11 June (later amended to noon on 10th) at the Shire Office Springfield. In a reference to the collapse of the King's Bridge, Sharland wrote: "Very sorry that you have been put to so much worry at Bendigo, and hope you will pull out with flying colors".

On Saturday 8 June, M&A mailed a tender for £210-16-6 for the whole job, of which £85 was for the Monier culvert. Unfortunately they addressed it to Sharland at Lancefield, another locality for which he served as Engineer. He wrote to say that, as he had not been at Lancefield during post office hours from 8th to 10th, the postal official had not been able to hand him the envelope, and it had not reached him at Springfield until the morning of 11 June. By this time, Council had let the contract to William Dwyer of Riddell's Creek at £220. Sharland expressed his regrets, but consoled the partners with the thought that M&A would have plenty of work throughout Victoria replacing timber culverts with Monier "if everything goes well in this case".

M&A then sent a drawing showing only the gorge and the footings to Gummow Forrest & Co in Sydney, asking them "to be good enough to direct us as to the proper form of the curve and as to the necessary thicknesses at Crown and Springing, and the disposition of the grid". The height of the embankment would be 17'-6" (5.33 m) above the crown of the arch. Baltzer replied with semi-graphical computations and a profile of the arch. Like M&A he arrived at a rise of 5 feet for the span of 12 feet, but chose more conservative thicknesses of 4" (102 mm) for the crown and 6" (152 mm) for the springing (c.f. 2.5" and 5"). Gregory was now able to prepare drawings for the falsework.

In mid-June, Dwyer made contact with M&A, saying he would drop in to see them, but could not start work without official approval from the PWD. A month later, he wrote "You will no doubt think I am slow about my contract at Springfield but the Works Dept have not yet sent an approval of my tender. I am carting the sand at my own risk. I am sorry to say I did not ask enough for it as the road is very heavy. I have to use two horses to take a yard load [0.76 m³]. I will write you again as soon as I get word to go on. I suppose it is lying in the office in town. You might drop in and ask the cause of the delay if it would not be too much trouble." Sharland also wrote that he had heard nothing, but would be in town shortly and would call at M&A's office. Finally, on 23 July he was able to advise that approval had been received.

Dwyer then asked M&A to send cement and a mixing board to Bolinda siding. He chose this rather than Lancefield Junction because the latter was too far to permit more than one journey per day. He advised that W. Mullaby, of Bolinda, was willing to drive M&A staff from railway station to site and back any day, allowing them to return to Melbourne on the evening train. On 29 July, David Mitchell's company consigned 36 casks of Emu Cement from its Burnley works to Bolinda. RCMPC had 9.3 cwt (472 kg) of reinforcing rods dispatched.

By Friday 16 August the footings were completed and set. Monash noted in his diary: "Leave for Bolinda with Palmer. Home by 8.00". John Palmer was to be M&A's man on site. The culvert was to be cast in four lengths of roughly 15 feet each. With assistance from Dwyer, Palmer had the first section complete by 21st. On 22nd he reported to Monash "Sir, this is a very bad place to get our Cement down, or it would not have taken me so long to Turne Monier, yesterday Wednesday Mr Shallard paid a visit & Seamed Satesfied the first Length of Monier took 7 Barrals of Cement I shall try to Strike the Drum tomorrow & try to Get a Second length in by Saturday Night".

On the night of Sunday 25th, a flood struck the culvert. The centres for Section 2 were washed 300 yards downstream. (Destruction of falsework by flash floods was a constant hazard and happened on other projects.) Monday was spent recovering the timber. Work was held up by constant rain, and flood water covering the foundations, until Friday 30th. The centres for No.2 section were then reset, the grid placed, and a start made on concreting. By now, Palmer had two men assisting him. They worked to finish Section 2 on the Sunday morning, 1 September. Half the culvert was now complete. Palmer reported: "I shall try very hard to get another Length in next week we can do it if the wether will only hold up & give us a chance, this Gulley rises very fast as it has so many hills to drain so you see that everything is against us."

Tuesday 3 September was spent "waiting for length [No.2] to dry out". This is ominous, because it indicates that M&A still thought cement gained strength as it dried. In fact, the early drying would have halted the chemical reaction between cement and water necessary for the mortar to gain strength.

If kept moist, the simple mortar used (without chemical additives) would have gained about 2/3 to 3/4 of its potential strength within 7 days, and almost all within 28 days.

Sections 3 and 4 were completed by 14 September, and Monash visited the site on 17th, bringing his daughter Bertha as far as Bolinda. On 18th ill-fortune struck M&A again, when a flood in Bendigo Creek washed the centres from under the newly-turned Wade St Bridge. The arch had not gained sufficient strength to support itself, and collapsed.

Palmer removed the formwork from Section 4 of Scott's Creek culvert the next day and travelled to Melbourne. On Friday 20th, Sharland wrote to Monash: "Dwyer has just called on me ... and states a couple of cracks have made their appearance in the Monier work. One running about level with the haunches and the other on top of the arch ... Dwyer is nervous about the matter so I would communicate with him or inspect the work if you think the matter deserves attention." On the Monday, Dwyer wired: "Scotts bridge collapsed Saturday". (Only the first section was involved.) M&A told him to return the "frames and laggings" to the site.

Both Monash and Anderson travelled to Springfield on 25th to examine the damage. They may have suspected that the filling around the arch had not been properly balanced, because Monash prepared a calculation of the stresses caused by lop-sided loading. On 27th he drew up a "New Design".

The mishap was most inconvenient for the Partnership. Added to the partial collapse of King's Bridge and the destruction of Wade Street Bridge, it made the task of selling Monier arch bridges much more difficult, at a time when they were working hard to gain a contract for the proposed Grant St bridge in Ballarat. Sharland wrote: "You have my sympathy in the trouble experienced in the construction of the Scott's Creek Culvert and you can rely on me doing all I can to prevent the mishap becoming public. I have every confidence that your firm will satisfactorily carry out the work taken in hand".

Preparing to rebuild the entire culvert, Dwyer started hauling sand from Riddell's Creek rather than Bolinda. He had to hire carters for a whole day, but they could not quite fit in two loads a day from Bolinda. M&A consigned supplies of cement, sand, stone, and reinforcing rods. Alex Lynch arrived at site on 9 October and under his direction work progressed rapidly, the culvert being half finished by 18th. However, progress was then interrupted by slow delivery of materials from Bolinda partly due to heavy rain. M&A were not amused when Lynch ran out of cement, but they consigned a further four casks. On 2 November he reported: "I consider that the culvert is now fairly set, and is in a very good condition." The first section was rapidly "getting hard on the outside". (Another hint that the process was not understood.) Lynch had started filling the previous day, carefully following Monash's instruction to balance the levels either side of the arch, and expected to leave in a few days' time.

M&A's series of misfortunes was probably due to an initial lack of understanding in some quarters of the correct procedures for mixing, placing, and curing concrete to ensure good quality. However, their opponents took the opportunity to blame the Monier system itself, and question the quality of the cement. This was an extreme embarrassment for David Mitchell, M&A's chief supplier and financial backer, and for Richard Taylor, Manager of the Australian Portland Cement works at Fyansford. (Taylor supplied cement and generous credit to M&A until Mitchell insisted that they should buy it from him to the limit of his capacity.)

Mitchell's manager, chemist John Gibson, tested the strength of samples made from neat cement, and from cement mixed with Springfield sand to form a mortar. The results, conveyed to M&A at the end of November, showed the cement to have been "of excellent quality greatly exceeding in strength the requirements of any Monier specification". He argued therefore that the sand used was largely to blame. It contained an "undue proportion of loamy matter" which soaked up much of the mixing water, depriving the cement of the necessary water for the chemical reaction. He suspected that because of the awkwardness of the site, the mortar had been mixed with very little water and then carried down to the culvert, by which time it had dried out even further. As well as hindering the reaction, this would also have made it difficult for the workers to compact the mortar, another factor of critical importance to strength. (Samples of the cement were sent also to Richard Taylor. His report largely confirmed Gibson's results.)

The letter from David Mitchell continued: "It is more than unfortunate that two disasters have overtaken work carried out by you with my cement. Apart from the serious loss to yourselves there is little doubt that no matter what may have been the cause of failure, a reflection is cast upon the quality of materials used, more particularly the cement, and this is a most serious matter to me. However, be that as it may, I feel that these unfortunate occurrences are not solely due to negligence on your part and although I think that the Springfield failure might have been averted had closer attention been given to the character of the sand used I have pleasure, under the circumstances, in presenting you with the forty four (44) casks supplied to reinstate the work. Needless to say I do not do this in any other spirit that that of rendering some friendly assistance and whilst at the same time certain the cement had nothing whatever to do with the failures I feel that you have to some extent been the victims of unfortuitous circumstances."

M&A experienced some delay in obtaining payment for the completed contract. Being subcontractors, they were obliged to wait until Dwyer had completed the entire contract and received payment. He, however, was soldiering on in the face of unforseen difficulty and an inadequate bid. At the end of November he wrote: "Gentlemen, In reply to your note re Scott's Creek Culvert, I may state that I am not yet finished nor am I likely to be this side of Christmas as I have got a lot of solid rock to contend with all blasting so you know what that means It has been an unfortunate job for me from beginning to end ... we can almost cross with a dray now the culvert is O.K. Mr Sharland was here this week ..." At the end of December he explained: "You will no doubt think I have quite forgot you through not writing but I have only got finished it was a dambdable bit of a cutting. Mr Sharland was there and everything was finished to his satisfaction ... wishing you a merry new year I remain Yours faithfully." A cheque for £65-11-0 was sent on Dwyer's behalf on 12 February 1902.

Data.

Contract Details.

As was customary, M&A prepared a draft contract for the culvert and embankment and sent it to Sharland. It catered for the possibility that a general contractor might win the job and choose to carry out the patent Monier work under M&A's direction. It specifies the Monier mortar as 3.5 parts sand to one of approved cement. Reinforcement is to be 1/4" (6.35 mm) rods or wire, with the main bars spaced at 3 to the foot, and the "distributing" bars at 4 to the foot. Work was to be "to the satisfaction of the patentees" - effectively M&A in Victoria. The contractor's price had to include all royalties. The Patentees would supply the Contractor with detailed instructions as to the carrying out of the arch, which must be strictly adhered to. Alternatively, the Contractor could sublet the whole of the Monier portion to M&A who would do this for £110 including royalties of £16/10/0. No filling was to be placed over any portion of the arch before it was 5 days old, and must then proceed at an equal height on each side of the arch until it was one foot above the crown. M&A's notes on the general conditions of contract (Shire of Springfield Contract No.9) record that a deposit of £10 was required, the job was to commence within 14 days and end in 14 weeks, with a penalty of £3 per week thereafter. The maintenance period was 10 weeks. The Engineer was to be sole Arbitrator in disputes. There is no mention of progress payments. Sketches show the embankment to be 24 ft (7.32 m) high at the creek, with slopes at 1.5 to 1.

Wages.

The Wages Sheet for the fortnight ended 24 October 1901 reads as follows, pay rates being expressed in shillings per day.
Wm Dwyer: 86 hrs @ 9/- = £4-16-9
R Harper labourer: 69 hrs @ 7/- = £3-0-0
J Dwyer horse and dray: 6 days h+d @ 12/- and 3.25 days labouring @ 7/- = 4-14-9
T Rountree horse and dray: 4 days + 19 hrs = 3-4-6
M Ryan, extra hand at Monier work: 10 hrs @ 7/- = 0-8-9
Alex Lynch. Board for two weeks: 18/-.

Computations.

JM's computations for the "new design" were similar to those shown elsewhere on this website link. They revealed relatively small bending stresses of plus or minus 45 psi. For full design loading he calculated stresses of 200 psi; for partial loading, 95 psi; and with fill 2 feet over the crown, 50 psi. Given the thinness of the shell and the probability of distortion of the arch profile during construction, it is unlikely that these figures give an accurate picture of the actual stresses.

Analyses of sand.

Gibson's results show that, although washing the sand made little difference to the strength of the mortar at 28 days (about 200 psi), there was a significant difference at 7 days (115 psi with unwashed sand; 155 with washed). Mixed with South Melbourne sand (unwashed) the 7 and 28 day strengths were 202 and 255. The Springfield sand looked satisfactory when dry, but washing revealed an undue proportion of loamy matter, including very fine particles.

An undated note initialled H.G.W gives the composition of sand from Riddell's Creek (air dried) as:

SiO289.9%
Fe2O3, Al2O37.78
MgO0.24
CaO0.20
Volatile Matters:0.85
Water:1.13
Sulphur, Arsenic:absent.