REGRESSIONS IN PORTLAND CEMENT CLINKER MICROSCOPY – CLINKER FROM 1906: A SNAPSHOT OF THE EARLY AMERICAN PORTLAND CEMENT AND ROAD

CONSTRUCTION INDUSTRIES

Karl Peterson, University of Toronto, Department of Civil Engineering, 35 Saint George Street, Toronto, ON M5S 1A4, CANADA, email: karl.peterson@utoronto.ca

Lawrence Mailloux, Miami Dade College, Department of Chemistry, Physics, & Earth Sciences, 11011 SW 104th Street, Miami, FL 33176-3393, USA, email: lmaillou@mdc.edu

Lawrence Sutter, Michigan Technological University, Michigan Tech Transportation Institute, 1400 Townsend Drive, Houghton , MI 49931-1295, USA, email: llsutter@mtu.edu

Abstract

Attendees, Twitter followers, and Facebook fans of the International Cement Microscopy Association are all, to varying degrees, admirers of vintage concrete and mortar. Starting with the pyramids and ending with whatever decade satisfies the definition of vintage, we all enjoy the opportunity to study building materials from the past and to speculate as to the specifics of their manufacture (regardless of the individual’s definition of what technically constitutes concrete or mortar). Furthermore, we are all admirers of Dr. Donald H. Campbell, to whom this symposium is dedicated. His beautiful visions of portland cement clinker have inspired many people, practitioners and non-practitioners alike, to appreciate the imaginatively named alite, belite, celite, and felite phases on a deeper level (not to mention the elusive delite, a lesser known phase which shall be the subject of a later paper). This paper describes an early instance of Blomeco’s Granitoid Concrete Blocked Pavement, a material that satisfies several of the aforementioned categories, in that it consists of both a mortar layer and a concrete layer, and thanks to a very low water to cement ratio and to a very coarsely ground portland cement, provides abundant pristine examples of clinker from 1906.

Introduction

In the western hemisphere, particularly in the State of Michigan, home of the American Concrete Institute, there are considerable bragging rights associated with the carefully-worded designation of the “World’s First Mile of Concrete Highway” a title frequently attributed to a portion of Woodward Avenue in Detroit, Wayne County, built in 1909 [1,2]. This section of pavement was later removed in 1922 to make way for a new highway. Upon closer scrutiny of the historic details related to this stretch of road (and upon closer scrutiny of the subtle “stretch” in the use of the term highway) the field of candidates eligible for this coveted title is considerably broadened. One such candidate is a concrete pavement, still intact, on the aptly named Portland Street of Calumet Michigan [3,4]. In 1906, the village of Red Jacket, Michigan (present day village of Calumet) constructed over 2.5 km (1½ miles) of street with Granitoid Concrete Blocked Pavement [5]. Figure 1 is from an advertising brochure extolling the virtues of Granitoid pavement, and includes a photograph of the project in Calumet [6]. Only 350 meters of the pavement are exposed at the surface today, mostly near the intersection of Portland and 7th streets. In celebration of one hundred years of service, in 2006 the Village of Calumet allowed a 150 mm (6 inch) diameter core to be retrieved from this pavement. A picture of the core and a scan of a slab cut from the core are shown in Figure 2.

Figure 1: Pages from 1910 advertising brochure with a view of Granitoid concrete pavement facing south down 6th Street, Calumet, Michigan at the intersection with Elm Street [6]. The large building to the east, an opera house, still operates today as the Calumet Theater.

Figure 2: Core from the pavement (left) and scanned slab cut from the core (right), with tic marks every inch on left-hand side, and every cm on right-hand side.

Construction and materials

As chronicled in the paper of the day, the Copper Country Evening News (CCEN) the winners of the bid to pave the streets were the R. S. Blome Company from Chicago, and J. J. Byers, a firm from the nearby city of Houghton, Michigan [7]. Blomeco were specialists in what by today’s standards would be considered a concrete pavement, while J. J. Byers specialized in creosoted wood block pavement. Another major contender, in terms of materials, was brick; a material eliminated from the running by a signed petition of resident tax-payers who felt it was too noisy – a common theme still exploited today in the epic struggle between asphalt and portland cement based concrete pavement [8].

The CCEN maintained a lively account of the project spanning eight months from bid opening to project completion. The novelty of a concrete road was captured in headlines with such descriptions as “Like Sidewalk Construction – Pavement, However, Is Marked Off Into Small Blocks” or  such explanations as “Rain Affects it – Portion of Sixth Street Pavement to Be Relaid at Once” [9,10]. Probably the most notable development was a situation where wooden blocks placed on 8th Street began popping out of place during construction [11,12]. Investigations and debate ensued, with experts brought in from as far away as Minneapolis, Minnesota to advise the Village Council [13,14]. Eventually, Granitoid persevered, with J. J. Byers ceding the remainder of the construction work to Blomeco [15].

In spite of the Granitoid namesake, excerpts from the Village Construction Specifications for the “Mixing and Laying of Concrete and Formation of Granolithic Block” were written loosely enough to preclude the use of granitic aggregate [16]. The specifications required only a wearing surface “consisting of one part of Portland cement and one and one half parts of granite screenings or clean sharp sand.” Not surprisingly, waste rock from the local burgeoning copper

mining industry was used as a substitute for granite (specifically Precambrian volcanic rocks dated at about 1100 Ma) [17]. As shown in Figure 3, flecks of native copper can even be observed in the aggregate on slabs cut from the core.

Figure 3: Native copper exposed at surface of slab cut from pavement core.

The pavement was constructed in two layers, the first was specified at a thickness 5 ¼ inches after compaction (the actual thickness of the lower layer in the core was measured at 4 ½ inches). The process of compaction was one of “ramming” the concrete into place, similar in many respects to modern roller-compacted concrete. According to the specifications, the concrete was to be made from a mixture of:

“one part of cement, three parts of sand, and four parts of crushed stone. The sand and cement shall be thoroughly mixed, dry, to which sufficient water shall be added and then made into a stiff mortar. The crushed stone shall then be immediately incorporated in the mortar and the mass thoroughly mixed adding water from time to time as the mixing progresses, until each particle of stone is covered with mortar.”

Accounts in the CCEN vary, sometimes the first concrete layer was allowed to cure over a period of a couple of days, and in other cases only a few hours before being covered with the second layer of mortar [5]. Although Village Specifications dictated that “the concrete shall be mixed on movable tight platforms” or mixing boards, mention is made in CCEN of a mixing machine [5]. No mention is made anywhere of the stamping machine pictured in Figure 4, and patented in 1910 by Rudolph S. Blome for “providing on the top surfaces of pavements, before they have set or hardened, impressions whereby to give the pavement a block-like appearance and to enable the draft animals to obtain more effective and secure footholds” [18]. However, the regularity of the pattern at the site suggests that a stamping device was likely employed. Village Specifications were also concerned with horse traffic, and stated that the brick shaped pattern:

Figure 4: Drawing of stamping machine for producing brick pattern on pavement surface [18].

“shall be formed with a grooving bar especially made for the purpose and when the pavement is complete the grooves shall be from one quarter inch to three-eighths inch in depth and shall present a slightly rounded upper edge so as to provide a firm and substantial foot-hold for horses.”

The only clue as to the origin of the portland cement came from a CCEN article that mentioned the Alpha brand, which at the time was manufactured at two sites near Martin’s Creek, New Jersey, in the Lehigh Valley cement manufacturing district [9,19,20]. Figure 5 shows an image of one of the Martin’s Creek plants, and two bulk oxide reports for Alpha portland cement from the era are included in Table 1 [21,19]. As to the veracity of the CCEN article, it is interesting to note that the local newspaper coverage of the 2006 coring of Portland Street mistakenly reported that the lower layer of the pavement consisted of “what looked like six inches of red sandstone” (a statement possibly inspired by the observation of red coring slurry, its color imparted by the abundant hematite present in the basalt) [22]. To investigate CCEN’s reporting, an effort to quantify the bulk oxide composition of the clinker by Scanning Electron Microscopy X-ray Energy Dispersive Spectroscopy (SEM-EDS) was initiated, but soon abandoned, due to the impractical nature such a strategy; particularly when posed with such questions as:

• Would the large remnant cement particles be representative of the cement as a whole? The answer: probably not.

• Are the differences in the reported bulk oxide compositions from portland cements from the era sufficient to distinguish one plant from another? Again, the answer: probably not.

• Would a trace element approach comparing cements produced from different regions be a better indicator? The answer: definitely yes, but beyond the resolution of the available EDS instrumentation.

Figure 5: Portion of page from the “Illustrated Story of a Bag of Alpha Cement” [21].

Table 1: Bulk oxide reports for Alpha portland cement [19], and bulk oxides recalculated from volume fractions observed in remnant cement grains based on ideal cement phase compositions.

wt. % oxide from Eckel, 1905 from Eckel, 1905 from Granitoid SiO2 22.89 22.62 21.8 Al2O3 8.00 8.76 8.0 Fe2O3 2.44 2.66 5.4 CaO 63.38 61.46 63.4 MgO 2.30 2.92 1.0

K2O+Na2O - - 0.2 SO3 - 1.53 0.2

CO2+H2O 0.99 - - sum 100.00 99.95 100

Clinker microscopy

In spite of the short-comings of the SEM-EDS approach, a collection of twenty-eight clinker particles were located and documented with back-scattered electron (BSE) images as shown in Figure 6. Characteristic x-ray elemental maps were collected from ten of these particles, and used to produce maps of the clinker phases as shown in Figures 7a and 7b. Representative spectra for the various phases have recently been collected, but the quantified EDS results were not calculated in time for inclusion in this paper.

Phase identification for the maps of Figure 7 was accomplished with Multi-Spec (a freeware multispectral image data analysis system) based on user-identified training sets for alite, belite, C3A, C4AF, periclase, hydration products, pore space, and potassium-sulfate phases [23,24]. The results found for the volume fractions of the clinker phases are summarized in Table 2. Assuming ideal compositions for the clinker phases, the equivalent bulk oxide composition was computed based on the volume fractions using specific gravity values from literature, and the result included in Table 1 [25].

Table 2: Summary of phase volume percentages from the ten phase maps, and conversion to phase weight percentages.

clinker phase vol. % spg wt. % M 4.9 3.581 5.5

C3S 29.1 3.153 28.8 C2S 49.2 3.143 48.5 C3A 10.1 3.030 9.6

C4AF 6.2 3.708 7.2 KS 0.5 2.668 0.4

Figure 6: Assortment of remnant cement grains. The magnification varies, but each frame is on the order of 0.1 to 0.3 mm wide.

Figure 7a: BSE images and results of phase classification.

Figure 7a: BSE images and results of phase classification.

Discussion and Conclusions

The comparison between bulk oxide results based on the SEM-EDS approach and bulk oxide results from the literature are in general agreement, with the SEM-EDS approach overestimating the amount of iron and underestimating the amount of magnesium. Part of the disparity may be due to the use of ideal clinker chemical compositions during the conversion of volume percentages to oxide weight percentages. For example, magnesium present in the ferrite phases was not included, thereby underestimating magnesium, and overestimating iron. As mentioned earlier, representative spectra have been collected from the clinker phases, but quantification of the spectra was not completed in time for inclusion in this paper. It is expected that the use of actual chemical compositions as opposed to ideal chemical compositions will yield a better approximation of the bulk oxides.

Another source of error in the process involved the classification of the elemental maps into clinker phases. A close examination of the phase maps indicates some confusion, especially along boundaries between phases. This was particularly a problem for the aluminate and ferrite phases, as they were often intermingled at a scale difficult to discern at the magnification used to produce the maps. In some cases there were issues with the distinction between alite and belite, resulting in belite crystals that appeared to be speckled or lined with alite. Some of this noise was removed through the use of a digital median filter, where isolated pixels of one phase are re-classified according to the majority of surrounding pixels of another phase, but the approach was not always successful.

Finally, it was not clear whether the historic bulk oxide accounts are for the clinker phases alone, or for the clinker and calcium sulfate phases together. Since the SEM-EDS approach examined only clinker particles, sources of calcium sulfate were not included in the bulk oxide composition.

References

1. Historical Site P25102, Informational Designation “First Mile of Concrete Highway,” Michigan State Housing Development Authority, Historic Preservation Office, 1957. http://www.mcgi.state.mi.us/hso/sites/15527.htm

2. “Again – Michigan’s Oldest Concrete Pavement? More on Michigan’s Oldest Concrete Paving Project,” Michigan Roads and Construction, Vol. 50, No. 29, p. 6, 1953.

3. Historical Site P23286, Informational Designation “First Use of Concrete Paving,” Michigan State Housing Development Authority, Historic Preservation Office, 1957. http://www.mcgi.state.mi.us/hso/sites/6012.htm

4. “Calumet Celebrates 50-year-old Concrete Pavements, Mark Calumet Paving Anniversary,” Michigan Roads and Construction, Vol. 53, No. 35, p. 2, 1956.

5. Mailloux, L., Peterson, K., Van Dam, T., Ellis, K. "Lessons from Michigan's Oldest Concrete Pavement - Still Serving after 100 Years," Proceedings of the 9th International Conference on Concrete Pavements, San Francisco, California, USA, August 17-21, 2008.

6. Which Pavement Shall We Select? Rudolph S. Blome Company, Chicago, DeGolyer Library Ephemera Collection, Trade Catalog #506, Southern Methodist University, Dallas, Texas, USA, circa 1910.

7. “Action Pleases – Selections of Paving Materials by Council Favored – Granitoid, Creosoted Wood - Believed Council Has Acted Wisely and Well and that Materials to Be Used Are the Best – Blome and Byers Get Contracts,” Copper Country Evening News, February 28th, p. 8, 1906.

8. “Much Interest – Opening of the Paving Bids Tonight Arouses Curiosity - Various Kinds of Materials - All Are Represented by Representatives of Concerns Desirous of Landing Contract – 6th Street Residents May Present Their Petition Against Brick,” Copper Country Evening News, February 15th, p. 6, 1906.

9. “As to Granitoid – Paving Material for 7th Street Will Be Made Here – Like Sidewalk Construction – Pavement, However, Is Marked Off Into Small Blocks, Making Firm Foothold for Horses,” Copper Country Evening News, May 1st, p. 8, 1906.

10. “Rain Affects It – Portion of 6th Street Pavement to Be Relaid at Once – May Finish 6th this Week – At Present Rate Men Will Be Enabled to Start Operations on Portland Street Next Week – Portion of Oak Street Creosoted Pavement Will Be Relaid,” Copper Country Evening News, September 5th, p. 6, 1906.

11. “Barnard Is Here – Creosoted Wood Pavement Man Talks on Houghton Pavement – Sand Was Swept off too Soon – Shows Letter from Government Inspector in New Orleans Which Shows Blocks Have Been in Use 35 Years,” Copper Country Evening News, June 25th, p. 8,1906.

12. “Report Denied – Rumor Request Would Be Made to Pave 6th Street with Brick – Will Use Creosoted Blocks – Rumor Gained Currency from Fact Blocks Had Sprung on 8th – An Ideal Pavement When Properly Cared For,” Copper Country Evening News, July 12th, p. 8, 1906.

13. “Two Engineers – Will Confer With the Council as to Creosote – May Use Another Material – Engineer Dutton of Minneapolis Sent for – Engineer Beehler Here – Conference Will Be Held,” Copper Country Evening News, July 19th, p. 8, 1906.

14. “Shumaker is Emphatic – Best Interests of Village to be Served in Paving Trouble – Beehler Also Talks – Says Whatever Ails Pavement is Due to a Local Condition – Talk of Using Granitoid for Creosoted Wood if Adverse Reports Are Made, Dutton Here Tomorrow to Make Investigations,” Copper Country Evening News, July 20th, p. 8, 1906.

15. “Granitoid on 6th Street – Substituted for Creosoted Wood, the Original Material – Meeting Held this Afternoon – James J. Byers Gives Up His Contract and the New Work is Awarded to Blome & Company – Village Will Save Several Thousand Dollars,” Copper Country Evening News, July 24th, p. 8, 1906.

16. Dorr, J. A., and D. F. Eschman, Geology of Michigan, The University of Michigan Press, Ann Arbor, Michigan, 1970.

17. Specifications for the Improvement of Streets for the Village of Red Jacket, Michigan, G. L. Clausen, Village Engineer, 1906.

18. United States Patent 967,714 Pavement-Blocking Device, R. S. Blome and W. J. Sinek, August 16, 1910.

19. Front, C. M., J. M. Christopher, and M. C. Fox. Images of America – The Lehigh Valley Cement Industry, Arcadia Publishing, Chicago, 2006.

20. Eckel E. C., Cements, Limes, and Plasters; Their Materials, Manufacture, and Properties, J. Wiley & Sons; New York, 1905. http://catalog.hathitrust.org/Record/005737488

21. Alpha, the Guaranteed Portland Cement : How to Use it, Alpha Portland Cement Co., Easton, PA, USA, 1917. http://catalog.hathitrust.org/Record/008897899

22. Nordberg, J. “Oldest Pavement Studied in Search for Concrete Truth,” Daily Mining Gazette, March 29th, p. 1, 2006.

23. Stutzman, P., “Scanning electron microscopy imaging of hydraulic cement microstructure,” Cement and Concrete Composites, Vol. 26, No. 8, p. 957-966, 2004.

24. Landgrebe, D, Biehl, L., MultiSpec© A freeware multispectral image data analysis system, Purdue Univeristy, West Lafayette, Indiana, USA, 2/20/2012. https://engineering.purdue.edu/~biehl/MultiSpec/

25. Balonis, M., Glasser, F.P. “The Density of Cement Phases,” Cement and Concrete Research, Vol. 39, No. 9, p. 733-739, 2009.