6-pdrs from Liège – A walk-around

Earlier I introduced this fine bronze cannon and it’s peculiar origin:

Vacation23 102

This cannon represents the 8th Indiana Battery at Chickamauga.  Let’s do a formal walk around this Belgian cannon, noting the features and markings.  As mentioned in the earlier post, the intent was to use this cannon to test European metal within an American pattern.  So the exterior form is very much “American”:

Vacation23 098

We see a slight muzzle swell, chase ring, a single reinforce extending just past the trunnions, a base ring, then a simple cascabel at the breech.  One could easily mistake this cannon for a 6-pdr Model 1841 Field Gun.

Looking close to the muzzle:

Vacation23 099

From the face, there is a simple cavetto and a fillet to join with the muzzle swell.  The chase ring is a full astragal with fillets on each side.  No other adorning features.

The trunnions themselves are simple forms also.  That on the right has no markings:

Vacation23 094
Turing to the left side, there are markings to interpret:

Vacation23 093

“No. 1 // 369 K”  The top line indicates this was the first of the set.  The second line may indicate the weight – 369 kg = 813.5 pounds.  Perhaps?  But that would put the 6-pdr about 65 pounds lighter than a Model 1841 of the same caliber. But close to the weight of a Model 1840 6-pdr.

Or 369 K might refer to a foundry sequence number?  There’s more markings to consider.

Normally on American weapons of this period, we’d find the weight stamped on the breech face, below the knob:

Vacation23 092

A little hard to read, but this stamp is 889.  That number, in pounds, conforms to the weight of a standard Model 1841 6-pdr.  Keep both these stampings (left trunnion and breech face) in mind when we look at the second Belgian gun.

We already discussed the nice label atop the base ring “Liège 1841”:

Vacation23 091

This is very much European in look.  Note also the vent, which is bouched.

We also looked at the muzzle in the earlier post:

Vacation23 095

No so much a “brag” here, as many guns would fire eight times that number during the Civil War.  “Fired.1000.Rounds. // 1842” was, I think, intended to mark this gun as the control example in testing.  After all, 1000 rounds was not a full life-cycle test.  That in mind, the bore diameter measures 93 mm, or 3.66142 inches. Giving allowances for my field measures and such, that’s very close to the standard 6-pdr bore specifications.

Moving to the other surviving example, on the north end of the Chickamauga battlefield, representing Douglas’s Texas Battery:

Vacation23 023

Again, a familiar exterior form.

This example has no markings on the muzzle face:

Vacation23 028

And about that muzzle, I measure it at 98 mm in diameter, or 3.85 inches.  And that is what we might expect for a cannon subjected to a great number of test fires.  The bore, however, is relatively smooth:

Vacation23 029

Looking aside from the mud-dauber nest and the bottle seated in the chamber, there’s not a lot of scratches.  Though the stress lines give the appearance of use.

The left trunnion has markings:

Vacation23 036

“No. 4 // 370 K”   The first, of course, indicates this was the last of the set purchased in Belgium.  And 370 kg would be 815.7 pounds.  Again, closer to the Model 1840, which would be the pattern in hand when the board was in Belgium, than the Model 1841.

But what about the breech face, is it stamped? Vacation23 035

Yes.  But it reads “1063”.   And that is NOT the weight of the piece.  So “might be” and “could be” on the stampings.

The base ring on No. 4 has the same label as the first gun:

Vacation23 040

However, note the vent.  This vent was cleared after what appears to be a lot of wear.

I didn’t have my full field kit in hand when visiting these guns.  So I didn’t get a chance to take overall measurements.  However, conveniently sited next to No. 4 is a standard Model 1841 6-pdr, registry number 146, cast by Ames in 1844:

Vacation23 048

The base ring sits directly atop the elevating screw.  The same as on the Belgian cannons.  These National Park Service reproduction carriages tend to be very uniform in dimensions.  So my first inclination is the Belgian guns match the Model 1841 in terms of length.  At least from the trunnions to base ring.

But if we go with the trunnion stamp as the weight in kilograms?  Hard to reconcile that with the dimensions.  One of these days I’ll return to the Belgian guns to give an exact number on the dimensions.  (Or perhaps a reader with a tape measure might save me the trip!)  And that might help discriminate the stampings.

Questions of weight and dimensions aside, there are two of the four Belgian bronze guns surviving – number 1 and number 4.   If we accept No. 1 as the “control” for experiments on endurance…. and No. 4, with all that bore wear, as subjected to substantial firings… then might we suppose No. 2 and No. 3 were destroyed in the process?

Though I’ve never run across a formal report of testing for these Belgian guns, I do know William Wade used the results in other experiments.  And for the next few years Wade would work to perfect the alloying process for bronze and other fine points of casting.  The result that played out on battlefields of the Mexican-American War.  Then later the Civil War.


Less successful sibling: The 13-inch Rodman Gun

I’ve detailed the design and production histories of the 8-inch, 10-inch, 12-inch Rifle, and 15-inch Rodman Guns.  Before I turn to the “big brother” of the set, the 20-inch Rodman, for sake of complete coverage I should mention the 13-inch Rodman smoothbore gun.

Thomas J. Rodman’s 1861 patterns included the 13-inch gun with proportions between that of the 10-inch and 15-inch guns:

Fort Pitt Foundry cast one 13-inch Rodman in mid-1863, the only gun of that caliber cast to Rodman’s original pattern.  Given the time frame, the gun likely had no preponderance and sockets for the updated elevating system.

Two more 13-inch guns appear in the records, but differ from the original pattern.  In 1866, Cyrus Alger delivered a pair of 13-inch Rodmans.  The first used the mold of a 15-inch Rodman, but bored to the smaller caliber, weighing over 51,000 pounds.  The exact pattern for Alger’s second is unknown, but it weighed 38,500 pounds.  The heavier of the pair went to the test range for trials and burst after firing over 700 rounds.  Testimony attributed the failure to tests of quick burning powder, which generated many times the pressure of the mammoth powder used in service.

Very likely the third 13-inch Rodman became the subject of a different experiment.  An 1878 report to Congress details the conversion of a 13-inch Rodman smoothbore to a 10-inch rifle.  Starting with a 13-inch gun cast in 1866, South Boston Foundry, which Alger’s facility became in the 1870s, reamed out the bore to 17 inches.  The foundry then installed a wrought iron rifled bore and jacket.  The rifling pattern included seventeen grooves each just under one inch.

The product weighed 40,320 pounds.  The weight is within the range expected if the second of Alger’s guns was the subject of conversion.  But the report does not offer a foundry or registry number for conformation.  Regardless the gun went to Sandy Hook, New York for tests.  Like many expedients of the late 19th century, the conversion offered little gain for the money expended.

While not produced in numbers, the 13-inch Rodman appeared in post-war plans.  In 1867 a board of engineer, artillery, and ordnance officers recommended the 13-, 15-, and 20-inch as the preferred smoothbore calibers.  The recommendation apparently carried weight, for as late as 1891 the Manual for Heavy Artillery listed 13-inch smoothbores as “in the system.”  Although General John C. Tidball noted it should be considered an “experimental type” as only a few were produced.  In some regards this reflected American strategic thought at the time.  Should war come again, the Army and Navy would have ample time to re-arm with the preferred weapons.

I would question why the 13-inch gun ever existed to begin with.  Certainly the 15-inch gun provided greater firepower.  In the context of coast defense, what part entrances could the 13-inch cover that the 10-inch could not?   Particularly with the coastal forts of the time positioned with the older seacoast guns in mind.  Then again I was not an ordnance officer in the 1860s!

None of the 13-inch guns survive today, leaving us with a few line drawings and congressional testimonies to tell their story.  And that story is but a footnote in the larger chapter of the Rodman guns.  The 13-inch guns were the unsuccessful sibling of the family.

15-inch Rodman Prototype: The Result of All Those Experiments

Having discussed the early casting experiments by Thomas J. Rodman to refine his casting technique and subsequent gunpowder tests, it is time I turned to the prototypes of the Rodman Guns.  Instead of constraining the next tests to just 10-inch weapons, the first prototype of the improved pattern increased to 15-inch.  The result was this gun – with registry number 1 – the prototype of a long line of Rodman guns.

The Lincoln Gun - Fort Pitt Foundry #1

Concurrent with the gunpowder experiments, Rodman “ran the numbers” to determine the optimum thickness of metal for a 15-inch gun.  I’m reluctant to walk through the reasoning here, due to space and my “physics for poets” background.  The short version, for those of us willing to neglect the particulars of square roots and other higher forms of math, is the gun needed a thickness of around 16 inches around the seat of the charge.  The thickness increased to 25 inches from the bore bottom to the exterior breech face.

Rodman also took the time to determine, based on his complied data, the best profile for the gun’s chamber (or bore bottom as he called it).  The preceding experiments indicated the traditional flat-bottomed bore tended to crack at the corners.  In the course of experiments Rodman found the sub-caliber chambers of the old Columbiads detrimental to performance.  He wrote, “There should be no angles, either salient or re-entrant, in the termination of the bore, but he surfaces of the bore and of its termination should be tangent along their lines of junction…. the semi-ellipsoid is believed to be the best and true termination.”  However, the casting plan diagram showed that of a hemisphere.

Casting Diagram and Plan of 15-inch Prototype

The diagram indicates the use of the Rodman neck vice the traditional cascabels.  The thin lines around the finished gun’s form indicate Rodman planned for a large amount of excess metal.  On surface examination, this resembles the casting technique developed by John Dahlgren for his naval guns.

Another point of similarity, which Dahlgren would later cite, is the exterior form.  Yes, the exteriors of Rodman and Dahlgren guns feature blended curves, lacking external fixtures.  But Dahlgren’s pre-war shell guns show the use of a pure cylinder over the chamber, traditionally the location of the first reinforce on the gun.  Rodman guns, as seen on the prototype plan, featured a near continuous curve expanding from the breech face to a maximum thickness over the seat of the charge, thence gradually tapering down to the muzzle.

The exterior form matched that called the “ordnance shape” and offered few right angles where stress would accumulate.  Rodman guns offered no “flats” save that of the muzzle face, rimbases, and trunnions.  If this form borrowed from Dahlgren, the paper trail has yet to be established.  On the other hand, simple examination of Army ordnance from 1841 through 1861, considering the columbiad trials and even the shape of the 1857 “Napoleon,” demonstrate the evolution to such blended curves.

In the fall of 1859, Rodman tested and selected the best iron from supplies available at Fort Pitt Foundry.   The foundry lit furnaces on December 23 to start the casting process.  As with previous hollow core castings, after pouring the molten iron, the foundrymen  poured water into the insert.  According to Rodman’s notes, water entered the insert at 36° and exited at 58°.  Rather low temperatures considering the last 10-inch casting initially had water entering at 80° and exiting at 102°, but in the warm month of August.  Notice the difference between entry and exit temperatures remained 22°, summer or winter.

After twenty-one hours, the water temperature exiting the insert dropped to 47°, and the workers removed the insert.  As with the earlier castings, water then poured directly into the interior bore.  Exit temperatures jumped to 86°.  Cooling continued for over 140 more hours.  All told the gun sat in the pit for 168 hours, far less than the time taken to cool a solid cast gun of half the size.

The finished gun weighed 49,099 pounds, with a 1200 pound preponderance at the breech.  It measured 190 inches long with a maximum diameter of 48.1 inches.  The gun measured 25 inches in diameter at the muzzle.  The 15 inch bore ran 156 inches deep.  Regardless of the metric, this was a large gun – indeed the largest produced up to that time in the United States.

In May 1860, the gun went to Fort Monroe for trials.  And how was a 25 ton cannon moved from Pittsburgh to Old Point Comfort?

For this purpose two strong trussed beams, 50 feet long, were prepared.  These beams were placed parallel to each other, and about 36 inches apart, their ends resting upon two bolsters placed transversely across the middle points of two 8-wheeled platform cars.  The gun was suspended under the two trussed beams, and between the cars; so that its weight was equally distributed over the 16 wheels of the two cars.

Thus packed, the gun moved on the Pennsylvania Central Railroad, Northern Central, and finally the Washington Branch.  In Washington, D.C. the gun, still on the cars rolled onto a heavy cargo vessel, which took the setup to Fort Monroe.

Once at Fort Monroe, Army personnel prepared the gun for proofing.  It was time to fire some large projectiles from one of the world’s largest guns.  I’ll discuss the testing and proofing of the 15-inch prototype next.


Thomas J. Rodman’s report of the 1859 experiments appears in Reports of Experiments on the Properties of Metals for Cannon, and the Qualities of Cannon Powder; with an Account of the Fabrication and Trial of a 15-inch Gun (Boston: Charles H. Crosby, 1861), pages 192-274, 281-282.

Origins of the Rodman – Early Experiments in the Rodman Method, Part 5

Other than perhaps the phenomenal descriptions of bursting guns flinging metal into the heavens, Thomas J. Rodman’s experiments were dry scientific procedures.  Rodman and other ordnance officers focused on gathering data to reduce the unknowns.  The experiments of 1849, 1851, 1856, and 1857 all built up to a conclusion from the experiments conducted at Pittsburgh in the fall of 1858.

In August 1858, Rodman gathered the iron left over from the 1857 experiment for casting into another pair of 10-inch Columbiads.  (Robert P. Parrott at West Point Foundry declined to cast another columbiad for this experiment.)  During the previous year, Fort Pitt Foundry suffered a major fire which destroyed substantial portions of the facilities.  However the pit and other structures related to the experiments were untouched.  Unclear is if the resumption of experiments was due to the fire, funding resumption, or simply personnel availability… or all three.  Regardless Rodman resumed testing guns almost a year to the day.

In addition to using iron from the 1857 experiment, the 1858 castings used the same pits and molds as the previous guns.  Casting started on August 11 with metal poured for foundry number 362 as hollow core and number 363 as solid cast.  Metal flowed into the molds for twelve minutes.

As with previous experiments, the hollow gun cooled first with an bore insert or core.  At initial casting, water entered the insert at 80° and left the casting at 102°.  After 16 hours, the water exited at 90°, and the foundrymen removed the insert.  Water then entered directly into the bore at 80° and existed at 144°.  Exit temperature gradually eased until the 81st hour it reached 81°, and observers deemed cooling complete.  The solid cast gun remained in its pit for an additional two days for cooling.

The hollow core gun contracted a third of an inch more than the solid cast gun at the breech during cooling.  And again Rodman noted the mottled appearance on the hollow cast gun compared to the dappled metal of the solid casting.  Otherwise both guns passed checks and were readied for firing.  Even though the guns used the same mold as the 1857 casting, internally they differed.  The 1858 casting eliminated the sub-caliber chamber, making these accurately “new Columbiad” models.

Vent biagram for 1858 tests showing bore and chamber arrangements

Firing commenced on October 2 with a solid shot and 20 pounds of powder.  The second fire consisted of a shell propelled by 24 pounds of powder.  At that point, shots used the standard 14 pound service charge with shot.  Fifty-four fires at different intervals increased the powder charge, ranging from 15 to 18 pounds. Otherwise the remainder of 2450 fires, completed on Christmas Eve, used standard service charges.  At the completion of these fires, both guns remained intact.  The solid cast gun exhibited cracks, but the hollow cast gun was nearly perfect.

Not content, the Ordnance Department ordered the guns to Fort Monroe, Virginia for more tests.  There Captain Alexander B. Dyer supervised 1632 more firings of each gun with 18 pound charges and sold shot.  Both guns remained intact, although the solid cast gun had significant bore erosion.   All told, both guns survived over 4080 fires.  This incredible endurance provided a wealth of data.

Rodman offered several observations in his report with the intent to insulate chance or accident from the test results.  At length, he demonstrated that the powder used for the 1858 tests had no impact on the results.  He further demonstrated the performance of the guns from both years proved the theories of metal resistance to pressure.

Regarding bore enlargement, the data reinforced Rodman’s expectations of the compressibility of metal.  The solid cast number 363 had significant bore enlargement just in front of the location the shot seated.  At some intervals of the testing, these enlargements were three times that of the hollow cast number 362.  Rodman attributed the enlargement to “stretching of the metal” at the point where the shot was at rest.  The force of the gasses produced by ignition of powder was thus greatest.  This effect was, in Rodman’s opinion, predictable based on knowledge gained of the metal’s properties – quality of metal and cooling technique.  This provided insight into a gun’s potential endurance.  “With these facts before us, we should be no longer astonished at the result of the powder proof of these guns.”

An interesting side note within the report of testing involved the cascabel performance.  Normally the complaint about these fixtures involved breakage during gun handling.  In the 1858 tests, both guns lost their cascabels during firing.  They just fell off!  Rodman attributed the cracks leading to cascabel failure to “accidental cause” rather than deterioration.  But the incident is cited as prompting him to redesign the cascabel to the form seen on later Rodman Guns.

In his recompilation, Rodman assessed all variables in comparing previous test results.  He eliminated all possible factors to account for the differences between the 1857 and 1858 guns, save the chamber arrangements.  Rodman concluded simpler interior arrangements were stronger.

Rodman demonstrated with the compilation of data that hollow-core, water-cooled guns demonstrated their superiority through the tests.  Only after careful controls of all variables could a solid cast gun (the last example cast) match the hollow-core guns in endurance – and even then showing hints of future failure due to bore enlargement and cracks.  The hollow-cast guns outperformed the solid cast guns terms of shots fired – in some cases up to a factor of ten.

Further, Rodman complained that private foundries, with limited supervision, could not provide consistent results.  Too many variables were in play and too many differences of opinion remained.  Rodman specifically pointed out that Robert Parrott felt the metal used in the 1857-8 experiments to be too heavy for use.  He complained, “… we are at present far from possessing a practical knowledge of cast iron in its application to gun foundering; and it is too much to expect of private enterprise to take up and prosecute so intricate and expensive an inquiry.”    In short while Rodman could explain gunfounding in abstract, West Point, South Boston, and other gunfounders could not apply those theories in practice without aid from the Ordnance Department.

However, Rodman expressed confidence in the overall results.  He was now ready to attempt a 15-inch gun constructed upon his principles, using the practices followed for the 1858 tests.  The technical history of the Rodman Gun now entered a new phase – prototyping.


Thomas J. Rodman’s report of the 1858 experiments appears in Reports of Experiments on the Properties of Metals for Cannon, and the Qualities of Cannon Powder; with an Account of the Fabrication and Trial of a 15-inch Gun (Boston: Charles H. Crosby, 1861), pages 101-128, 277-278.

Origins of the Rodman – Early Experiments in the Rodman Method, Part 4

Previous posts in this set have covered experiments of the Rodman Method conducted in 1849, 1851, and 1856.  I’ll turn next to the next set of experiments conducted in 1857 and 1858.  I know the tally of these experiments is somewhat mundane, but I’ve not seen these discussed at length outside Rodman’s official reports.  Most accounts point to decade long experiments, leaving the details below the surface.  In addition, my thesis looking through all this is the hollow-core, water-cooled method was only a part of the solution.  Further there is an “administrative” story ranging from Rodman’s patent award, royalty payments, Ordnance Department policies, and congressional inquiries that runs parallel to that of the gunmaking evolutions.

In 1857, Captain Thomas J. Rodman received permission to continue tests.  For this round of tests Rodman cast three 10-inch Columbiads.  As with earlier tests, Fort Pitt Foundry produced castings, one solid and one hollow.  Added to the sample set was a solid cast gun from West Point.  Working with Lieutenant F. I. Shunk, Rodman graded and selected iron for these castings from stocks at West Point.  Each casting consumed equal portions of the various gradings, to ensure near identical mix.

Rodman intended to cast all three guns on the same day, but “owing to mismanagement in the Post Office Department” was unable to coordinate operations.  Fort Pitt poured metal for its castings on August 13, 1857.  West Point cast its gun on the following day.  The West Point gun cooled in an enclosed pit, packed with “green” or moist sand.  The Fort Pitt guns cooled in open pits as with the previous experiment.   Of note, Rodman reported the Fort Pitt castings filled in eight (hollow) and nine and a half minutes (solid) of pouring molten metal.

The West Point gun, cast solid, received foundry number 983 and cooled within eight days. Likewise Fort Pitt gun removed its solid cast number 335 after eight days of cooling in the pit.  According to Rodman’s observations, the solid cast guns had a cloudy and somewhat dappled appearance.

Fort Pitt foundry number 334, cast hollow, cooled in the same process as earlier hollow core guns.  For the first sixteen hours of cooling, water circulated through the core insert.  Water entered at 75° and exited at 95°.  After removing the core, water circulated directly through the bore, entering at 75° and exited at 136°.  After 19 ½ hours, the crew increased water flow and exit temperatures gradually decreased.  After five days of water cooling, number 334 came out of the pit and the casting flask on August 21.  Rodman noted the hollow core gun’s metal had a “much finer, and more uniformly mottled appearance.”

In terms of pattern, fracture diagrams indicate both foundries used the Pattern of 1844, but with some minor changes.  The form omitted the base ring for instance.  So these are “transitional” guns to the “new Columbiad” pattern.

Fracture Diagram of Fort Pitt 335

These columbiads retained the sub-caliber chamber of the older Model 1844 pattern – a twelve inch deep, eight inch diameter chamber.

After the West Point gun arrived in Pittsburgh, all guns went to the proofing range on October 22.  The first fire of all guns was with a solid shot and 20 pounds of powder.  The second fire consisted of a shell propelled by 24 pounds of powder.  At that point, shots used the standard 14 pound service charge, but alternating shot and shell.  Test firings used fresh powder from 1857 batches directly from DuPont, described as uniform in grain and free of dust.

After 88 fires, both solid cast guns exhibited cracks in the second reinforces.  The West Point gun completely failed on the 169th fire.  The Fort Pitt solid gun survived until the 399th fire.  Upon examination of the fragments, the Fort Pitt gun exhibited some small cavities left from the casting, which likely aided the failure.  Although not referenced in the report from 1857, Rodman later noted the West Point gun was heavier, denser, and in general stronger in almost every possible measure.  Yet it failed first.

Fracture Diagram of West Point 983

But the hollow core gun continued to fire through tests unfazed until the 600th shot.  Firing continued as crews drilled a second then a third new vent.  Cracks continued to develop through 1000 and even 1600 fires.  But Rodman described these cracks as different, “not the tortuous appearance of those in the other guns, but had more the appearance of having been cut and burned out by the gas.”

In his report to the Ordnance Department, Rodman provided the usual detailed measurements of bore, chamber, and vent enlargements taken after each firing.  He also provided meteorological data for each day of firing.   Tests of the metal showed the hollow core number 334 between the two solid cast guns in terms of density and tenacity.  The West Point gun actually used more dense and stronger metal.

Of the 1857 tests, Rodman offered very few conclusions.  The pattern of fracture and cracks did provide support for Rodman’s theories about gas expansion within the gun’s chambers. Concurrent with the big gun test of 1856 and 1857, Rodman also conducted experiments on a smaller scale.  Some of these focused on the behavior of iron cylinders containing the force of powder ignition.

One of Many Diagrams from Rodman's Cylinder Tests

Other tests assessed the metal’s resistance to force, or what Rodman called “potential for work.”  As Rodman noted, the performance of the gun had much to do with the elasticity and compressibility of the metal, perhaps more than with its strength.  More importantly, Rodman felt there was predictability to the performance of the metal.  Much of the data gathered from the 1857 tests provided support for a more substantial report the following year, which I’ll look at next.


The 1857 report of Thomas J. Rodman appears in Reports of Experiments on the Properties of Metals for Cannon, and the Qualities of Cannon Powder; with an Account of the Fabrication and Trial of a 15-inch Gun (Boston: Charles H. Crosby, 1861), pages 59-89.

Origins of the Rodman – Early Experiments in the Rodman Method, Part 3

As seen in earlier posts, William Wade of the Fort Pitt Foundry conducted experiments in 1849 and 1851 to proof the system proposed by Lieutenant Thomas J. Rodman.  While providing promising results, the tests did not provide conclusive proof that hollow-core, water-cooled casting produced the quality guns needed by the Army (and the Navy).

Indeed after the 1851 experiments, Rodman felt the need to elaborate on the test results in correspondence directly to the Chief of Ordnance, Colonel H. K. Craig. In November 1851, Rodman provided a detailed explanation for the difference in performance between the 8-inch and 10-inch hollow-core Columbiads tested earlier that year.  After extensive examination of cooling and force dynamics, Rodman explained this difference of performance by the rate of cooling.  This of course due to the sand around the 10-inch gun.  Boldly, Rodman predicted, “I have the utmost confidence that with iron of the same quality as that used in these guns, and a proper application of the new mode, a 10 inch gun may be made to endure 1000 to 1250 fires.”1

Rodman tested his theories again in 1852 using two 32-pdr guns (I overlooked this test in my earlier post), one hollow and one solid cast (foundry numbers 160 and 161 respectively).  The solid cast gun fired 1000 rounds of standard service charge of 8 pounds of powder, then burst after six shots with extreme charge of 10 ⅔ pounds.  The hollow gun likewise fired 1000 rounds using the standard charge, then twenty with the extreme charge.  The gun finally burst on the 1021st fire with two solid shot and 16 pounds of powder.  These tests offered no additional conclusions.2

Compilation of Rodman's Tests 1849-1858

Not until August 1856 did Rodman again have permission to test the heavy Columbiads.  Using carefully selected iron, Rodman directed the casting another matched pair, one hollow and one solid, 10-inch Columbiads, foundry numbers 331 and 332 respectively.  Fracture diagrams made after testing show the pattern followed that of Model 1844.

Casting started on August 23.  The hollow core gun followed the same cooling technique as the 1851 casting, but with the central core removed after 17 hours of cooling.  Rodman noted the use of a separate wrought iron tube sunk into the gunhead to allow the flow of water out of the empty bore at that point, preventing water from touching the hot exterior.

All proceeded fine until the 38th hour of cooling.  The pit attendants, acting without instructions, piled hot coals around the casting.  This kept the casting at a constant temperature for the next twenty hours.  Upon noticing this issue, the foundrymen cleared the coals, but the gun required over a hundred more hours to cool.  This allowed the gun to anneal, lowering the tenacity of metal.  Hollow cast number 331 came out of its casket on September 1.  The  solid cast number 332 came out on September 5 (providing some comparison of cooling times).

The hollow cast gun ran into more problems.  In cooling it contracted significantly more than the solid cast gun.  Worse, when taken to the lathe, the workers found the bore to be off axis, requiring careful machining to correct.  Not a good start for the gun.

On the test range, both guns went through proofing with a first shot with 20 pounds of powder and a solid shot.  Second fire was a shell with 24 pounds of powder.  All subsequent fires used solid shot and 18 pounds of powder.  The solid cast gun burst on the the 26th fire.  The hollow cast gun lasted a bit longer to 315 fires.  This was not the endurance Rodman hoped for.3

Examination of the metal after bursting indicated the hollow cast gun exhibited slightly better density and tenacity over the solid cast.  Aside from detailed examinations of bore and vent enlargements, Rodman focused on the fracture behavior. Taking into account the strain and likely expansion of the metal, Rodman provided formulas that not only predicted the area of failure but explained in part why internal cooling (hollow core) worked better.4

Predicted Bursting Patterns

Rodman also focused on the performance of the powder in the firings.  Complaining of the selection, Rodman felt it burned too quickly, stressing the gun without any performance gain. With this in mind, Rodman also pursued tests to determine the best grain and size of powder for Columbiads.5  While a tangent, this would bear fruit later in the development of the large guns.

In his conclusions, Rodman placed blame on the failed hollow core gun on the improper cooling rate.  This was a “throw away” test in that regard.  But remarkably, Rodman appears to have taken advantage of this to gather data for the next round of tests.  Closing his report, he listed a full page of unknown factors, including iron quality, casting technique, chamber dimensions, exterior form, and powder charge. And Rodman set a course to resolve those unknowns through further experiments, which I will continue with shortly.  But let me offer up for now Rodman’s closing justifying further experiments:

And it is believed that the true interests of the country would be promoted, in a military point of view, by entering, at as early a period as practicable, upon a series of experiments which would supply positive knowledge in place of probability in some, and positive ignorance in many other points of the utmost importance to the national defense; for it is better that millions should be expended in time of peace, and from an overflowing treasury, than that a single gun should burst in action.6



  1. Thomas J. Rodman,  “Report on the Causes of Difference in the Endurance of Cannon, Cast Solid and Cast Hollow and Cooled from the Exterior and Interior” dated November 30, 1851,  Reports of experiments on the strength and other properties of metals for Cannon (Philadelphia: Henry Carey Baird, 1856), pages 209-13.
  2. Details from chart found in Thomas J. Rodman, Reports of Experiments on the Properties of Metals for Cannon, and the Qualities of Cannon Powder; with an Account of the Fabrication and Trial of a 15-inch Gun (Boston: Charles H. Crosby, 1861), page 133.
  3. Ibid, pages 4-12.
  4. Ibid, pages 35-48.
  5. Ibid, page 15.
  6. Ibid, page 55.

Origins of the Rodman – Early Experiments in the Rodman Method, Part 2

In part one, I discussed two 8-inch Columbiads cast in 1849 and used to test the Rodman method of casting large guns.  The conclusion of that experiment indicated a slight improvement to the gun, but not great enough to prove the hollow-core, water-cooled method was greatly superior.  Gunfounder William Wade suggested more tests with better quality metal.

Lieutenant Thomas J. Rodman, working closely with Wade, waited until 1851 for the next full-scale tests.  Fort Pitt Foundry cast another pair of 8-inch Columbiads – one solid and one hollow-core – on July 30, 1851.   Then on August 21 of that year, Fort Pitt cast a similar pair of 10-inch Columbiads.  Wade selected what was considered high quality iron from Greenwood Furnace of New York.

Rodman revised the cooling technique after the first experiment.  For this 8-inch hollow-core columbiad, the gunmakers left the insert in the gun bore for the first twenty-five hours of cooling.  The insert (just as with the earlier experiment) consisted of a tube, closed at the bottom.  Inside that tube, a smaller diameter tube, open at the bottom, allowed water to flow down into the bore.  The water then ran off from the top of the casting mold into a trough.  All the while, a fire burned under the casting mold in the open pit to slow the overall cooling of the gun.

Diagram of Rodman Casting Arrangements

After that initial cooling phase, the foundrymen removed the core and let water flow directly into the now hardened gun bore.  Excess water continued to run out from the top of the casting mold into the trough.  This continued for forty hours.   About halfway through that period, the gunmakers extinguished the fire in the pit, allowing the gun to cool completely in the last twenty hours.  Wade indicated that for the 8-inch Columbiad 10,000 cubic feet (300 tons) of water circulated through the gun during the entire sixty-five hours of cooling.

As seen on the fracture diagrams supplied by Wade, the guns conformed to the Model 1844 pattern.

Fracture diagram of 8-inch solid cast Columbiad No. 3

Casting of the 10-inch hollow core Columbiad differed in both measures and technique.  Foundrymen left the core insert in the bore of the 10-inch Columbiad for ninety-four hours.  At that time extraction failed, as the bore had contracted around the core.  With the interior tube removed, water continued to flow at a reduced rate into the 10-inch Columbiad for an additional forty-eight hours.  Although Wade does not state such, presumably the core insert was drilled or wedged out after the final cooling cycle.

There was one other critical difference in the casting of the 10-inch hollow-core Columbiad.  Because the casting mold (both solid and hollow castings) for the 10-inch Columbiad were not large enough to allow what Wade considered sufficient clay, the foundrymen placed both 10-inch Columbiads in a pit filled with “green sand” to ensure the hot metal did not melt the casket of the casting mold.  This, as we shall see, changed the cooling rates and performance of the gun.

Fracture diagram of 10-inch solid cast Columbiad No. 5

With the guns cooled, all four went to the range.  The 8-inch Columbiads went through test firings between August 28 and October 2 of that year.  The 10-inch Columbiads followed in October 7 to October 18.  Just as with the 1849 tests, all four guns fired from a test mounting.  Gunners supplied proofing charges for the first two fires.  The 8-inchers fired a 12 pound charge for the first shot then a 15 pound charge for the second, after which the guns fired a 10 pound standard service charge.  Likewise the 10-inch Columbiads first fired a 20 pound charge, followed by a 24 pound, then settling on the 18 pound service charge.  The 8-inch Columbiads shots alternated between a 63 ½ pound solid shot or a 49 pound shell.  The 10-inch Columbiads fired 124 pound solid shot and 91 pound shells.

Wade reported a mixed bag of results:

  • 8-inch Columbiad, No. 3, cast solid – burst after 73 fires
  • 8-inch Columbiad, No. 4, cast hollow – survived 1500 fires
  • 10-inch Columbiad, No. 5, cast solid – burst after 20 fires
  • 10-inch Columbiad, No. 6, cast hollow – burst after 249 fires

The description of the bursting of 10-inch No.5 speaks to the power of black powder.  Wade wrote, “…the upper part, weighing 4400 pounds, was thrown upward, and fell in the rear, about 80 feet distant.  In its flight, it broke a limb of a tree, on a hill side, sixty feet above the level of the gun when fired.”

While the 8-inch No. 4 proved an excellent argument in favor of Rodman’s method, the 10-inch No. 6 dampened that success.  Upon examination of the fragments, Wade noted several cavities and fissures in the metal even describing some sections as “sponge-like”.  Wade attributed these flaws to the use of sand for exterior cooling, estimating that seven-tenths of the heat dissipated through the exterior as opposed to the water in the interior.  Most of these fissures appeared on the front part of the gun.  One large fissure was noted on the fracture diagram, along the chase of the Columbiad.

Fracture diagram of 10-inch hollow core Columbiad No.6

However, Wade insisted the fissures did not cause the gun’s failure.  As he noted, the crack in the gun started over the chamber and proceeded forward along the barrel, crossing over the fissure.  Instead, Wade attributed the failure of the gun to the brittle nature of the metal overall, a property caused by the rate of cooling of the gun.

Examination of the gun metal at the conclusion of these tests indicated, just as with the 1849 tests, no significant difference in the tensile strength of the metal of the different guns.  After assuring the paired castings took place under identical circumstances with the same batch of raw metal, Wade proclaimed, “The great difference of endurance must therefore be ascribed, to the different methods by which the castings were cooled; and to them alone.” The method, as Wade and Rodman would explain in detail, allowed the gun to contract upon itself while cooling, thus changing the way the metal reacted to the stress of firing.

In his report Wade made the conclusion that casting large caliber guns required even more fastidious selection of metal for gunmaking, calling for softer and weaker iron that underwent a managed cooling process.  He further added,

The method devised by Lieutenant Rodman, for accelerating the cooling of the interior of the guns, by passing a stream of water through them, and for retarding, at the same time, the cooling of the exterior, by surrounding it with heated air, appears to lead in the right direction, even if it does not fully accomplish the purpose designed.

Wade recommended further tests, with 32-pdr caliber guns in addition to the large caliber columbiads, with the aim to refine the process further.  (To my knowledge, the 32-pdr experiments were never conducted.)  CORRECTION: I overlooked the mention of tests in 1852.  Details in the next post in this series.

Rodman’s technique required even more refinement and some concurrent experiments with metal composition before it would bear fruit at the end of the 1850s.  As with many successful weapon systems throughout military history, the Rodman gun was not so much “revolutionary” but rather “evolutionary.”



William Wade.  “Report on the Manufacture and the Extreme Proof of two 8 inch Columbiads and of two 10 inch Columbiads….” dated January 24, 1852.  Reports of experiments on the strength and other properties of metals for Cannon. Philadelphia: Henry Carey Baird, 1856,  Pages 181-204.

Olmstead, Edwin, Wayne E. Stark, and Spencer C. Tucker. The Big Guns: Civil War Siege, Seacoast and Naval Cannon. Alexandria Bay, NY: Museum Restoration Service, 1997, Appendicies C115 and C142, pages 237 and 253.