Han Dynasty Crossbow III

Mar 2014
Folding and hammering was done by every culture that worked iron. Real question is how Chinese got high carbon iron, that we usually call steel. As far as I know Chinese were making cast iron very early, but could they make steel from cast iron? There are couple of lines in Wiki that there are some written source that mentions some kind of process. So, does anyone know if medieval China had ability to produce large amount of steel from cast iron?
Feb 2011
Don't take my word for it but from my understanding the mold for the cast iron would be partially destroyed (unlike bronze piece molding), so it would be much harder to mass produce it. You technically could, but it was be very hard to mass produce triggers with interchangeable parts when the mold keeps being destroyed or damaged after each casting. Again, I'm not sure about this.

From Yang Hong in Weapons of Ancient China on Han dynasty steel making (pg 185-186):

The pig iron from smelting furnaces or the scrap iron of used iron wares was used as the raw material for melting. Cupola furnaces were discovered at Wafangzhuang of City Nanyang, Tieshenggou of Gongxian, Guxingzhen of City Zhengzhou and Wangchenggan of Lushangxian. The adoption of the cupola furnace greatly improved the quality of the liquid molten iron, and therefore the quality of the articles cast. It is estimated that the average inner diameter of the cupola furnaces discovered at Wafangzhuong of City Nanyang was about 1.5 metres and their height about 3-4 metres. The operation of a furnace of this size is believed to have been semi-continuous. At the same time, the art of casting with clay, layered, and metal moulds was also perfected to a high degree.

The method of steel making and quenching techniques were also developed, and these were closely related with the manufacture of weapons. The low carbon steel made from pig iron during the Warring States Period developed into low-grade bai liangang (steel forged a 100 times) during the Western Han Dynasty. This was a significant breakthrough in steel making. Broad swords and double-edged swords made of this kind of low grade tempered steel were already in use by the end of the second century B.C. This has been proved by metallographic examination of the double-edged swords which were carried on the belt, other steel double-edged swords, and knives for cutting paper which are inlaid with gold. All of these were unearthed from the tomb of Emperor Liu Sheng at Mancheng, Hebei. The raw materials from which the steel of these swords was made was no different from that of the relics of the last years of the Warring States Period. Both are low carbon steel made from pig iron mixed with carbon, but the quality of the former is much better. Specifically, both the size and the number of impurities are reduced, the carbon content is also lower, and the various layers have a more homogeneous structure. The low-carbon layer of the double-edged steel swords unearthed from the secondary capital of state Yan is about 0.2 mm thick, and that of the double-edged steel swords unearthed from the tomb of Emperor Liu Sheng, about 0.05 – 1 mm. This reduction in the thickness of the layer is due to repeated forging, a process leading to bai lian gang (steel forged a hundred times).

The above findings show that the art of repeated heating and forging, adding solid-state carbon and hammering after repeated folding was known at that time and that a low-grade tempered steel was successfully produced. At the same time, following the development of the technique of annealing, the technology of making steel by decarbonizing cast iron through annealing was perfected. On the basis of this, therefore, a new technology using as open hearth process to decarbonize pig iron and turn it into low carbon steel while still in a semi-molten state developed. The steel made in this way is the so called “stir-fried steel” (chao gang). In the excavation of the site at Tieshenggou of Gongxian, a jar-shaped coverter furnace was discovered. According to the opinions of experts which are based on a study of the furnace, when air is blown into the furnace from the top, the shape of the furnace would ensure the oxidizing condition within it and accelerate the decarbonization therein. The capacity of this jar-shaped steel furnace is small, and it is easy to maintain the temperature in it due to its shape. Moreover, since it is built underground, the dissipation of heat is negligible and thus it is easy to raise the temperature in the furnace. The opening in the lower part is large, so it is easy to charge, to stir, and to let out the content of the furnace. The structure of the furnace is simple and it therefore is easy to build. These advantages facilitated its popularity at the time. Several such furnaces were found on the site of iron smelting at Wafangzhuang of City Nanyang. They are quite similar to the one at Tieshenggou of Gongxian, and some still have lumps of iron at the bottom. The advent of this new technique of decarbonization gave new impetus to the development of highly tempered steel (bai lian gang) and removed the barrier resulting from low steel production and inefficiency of pig iron blending with carbon. Thus it raised the production and quality of the steel and made it possible for highly tempered decarbonized steel (bai lian gang) to come into its own. Recently two steel weapons with inscriptions were discovered in Eastern Han tombs in Shandong and Jiangsu. One of them, a ringheaded steel sword was made in the year of the reign of the Eastern Han Dynasty (112 A.D.), was unearthed from Changshan or Shandong in 1974. An examination showed that the sword was made of chao gang (Fig. 304). The structure of the steel of the sword magnified 10,000 times under a microscope shows the sword has 30 layers. This agrees with the inscription on the sword which says the sword was forged 30 times, that is the steel was heated, folded and hammered 30 times.

The other weapons is a steel double-edged sword (jian) made in the second year of the reign of Jian Chu of Eastern Han (78 A.D.) and unearthed in Xuzhou in 1978 (Fig. 300). Its date of manufacture is a quarter of a century earlier than that of the iron broad sword unearthed at Cang Shan. It inscription says it was made by being forged 50 times under the instructions of the governor of the prefecture of Shu. Miscroscopic examination of the metal structure revealed that it was forged from high-carbon “stir-fried” steel (chao gang). These two specimens are concrete evidences that the technology of bai lian gang steel, i.e. highly tempered and refined steel, was completely mastered during the Eastern Han Dynasty. They also demonstrate that the technology was applied to the manufacturing of weapons.

The advancement of the technique of quenching is very important to the manufacturing of weapons. The steel broad sword and double-edged swords unearthed from the tomb of emperor Liu Sheng at mancheng, show that the technique of partial quenching, that is, quenching the cutting edge to increase its hardness while leaving the toughness of the other parts of the weapon such as the back, central ridge untouched, was already mastered at the time. This technique increased the quality and effectiveness of weapons greatly. At the same time, methods of decarbonization and annealing to increase the malleability of the steel and to make it more pliable for processing, was also used.

And now for some tables:

Above two tables came from: Sarah Taylor, Early Chinese Iron Technology: Some Social And Historical Implications

Anyway, I would appreciate it if we focus on crossbows. I might make a separate thread about Chinese metallurgy in the future.
Last edited:
Jun 2019
Southeast Asia
Thank you, but I still think that they could cast it with the bronze casting method.

Anyway, you are right that we should focus on the crossbow.

How do you aim with the Han crossbow?

I heard that there are Chinese crossbows lock from the 650 BC. What did the Han crossbow improve over the previous model?
Feb 2011
You can read about aiming in post 6 of this thread.

The main form of a crossbow lock is basically to have a trigger guard. It's not a "true" lock but it still helps prevent the trigger from being accidentally activated.

Qin bronze crossbow:

^You can see there's a trigger guard for that too.

The Mawangdui crossbow lacks a trigger guard:

Atarn showed a crossbow trigger that took it one step further, here there's a trigger lock that prevents the trigger from being activated:

And here the trigger lock is 'unlocked', so to speak:

The trigger pull for Chinese crossbow triggers was incredibly light, so there was a pretty big incentive for them to have some sort of guard.
Last edited:
Jun 2019
Southeast Asia
In comparison with European crossbow, I noticed that the one of the earliest depiction of European crossbow have vertical trigger.

Weapon Europe 10th century 989-1010 Commentaries of Hayman on Ezekiel, Auxerre, France 5.jpg

It seems that European crossbow trigger reach its horizontal form during the 11th century.

1050-1100 Spanish manuscript

Weapon Europe 11th century 1050-1100 Spain Roda Bible crossbow.jpg

1086 Spanish manuscript

Weapon Europe 11th century 1086 crossbow Spain Burgo de Osma copy.jpg

It certainly look more compact as it did not have a long stock, but it probably have shorter power stroke compared to later European crossbow. Did the European miss the chance to make a compact crossbow by not improving on vertical trigger design?

How similar are Han Dynasty mechanism to later European pistol crossbow trigger?

Weapon Europe 19th century Pistol crossbow 1.jpg

Also one thing that bother me, is it necessary to create complex trigger design for crossbow like the one used in Europe and China?

Because it seems that the various crossbow used by tribes in China can also release their bolts without using such designs.

The Crossbows of South-west China

I also think that based on the pinching method shown in Timurid, Safavid and Mughal crossbow, that it may be possible to create a very simple trigger based on pushing by the thumb.

Crossbow Timurid probable pinching trigger design.png

While my picture may not be efficient enough, I find these more efficient pinch lever design.

early all-wooden hunting crossbow (rising peg trigger).jpg

There is nothing preventing the maker from putting trigger guard in this design of either hook guard or the blocking design.
Last edited:
Feb 2011
A vertical trigger in itself doesn't make the trigger compact, it's about the size of the trigger. It's technically possible for the Medieval European trigger to adjust itself to have a vertical trigger as long as if you're OK with an abnormally long grip to the point of being unergonomic, because the trigger lever still must be very long due to the high trigger pull. It'll end up looking something like this:

The higher the draw weight the longer the lever has to be. Albeit the above has a sear which reduces the need for a long lever, but this type of design only appeared in very late Medieval Europe. For Medieval crossbow triggers without that sear, you could make a short trigger lever but the draw weight would need to be light in order to offset the increased trigger pull. European crossbow design started improving on the trigger pull (and hence shorter trigger lever for the same draw weight) by the late 15th century Medieval period.
And frankly that first picture you shared seem to have taken artistic license as a vertical trigger requires a grip and the picture simply don't show it. The earliest description of Medieval European crossbows were probably of similar power to Han crossbows, except much much larger due in no small part to the size of the trigger lever:

This cross-bow is a bow of the barbarians quite unknown to the Greeks; and it is not stretched by the right hand pulling the string whilst the left pulls the bow in a contrary direction, but he who stretches this warlike and very far-shooting weapon must lie, one might say, almost on his back and apply both feet strongly against the semi-circle of the bow and with his two hands pull the string with all his might in the contrary direction. In the middle of the string is a socket, a cylindrical kind of cup fitted to the string itself, and about as long as an arrow of considerable size which reaches from the string to the very middle of the bow; and through this arrows of many sorts are shot out. The arrows used with this bow are very short in length, but very thick, fitted in front with a very heavy iron tip. And in discharging them the string shoots them out with enormous violence and force, and whatever these darts chance to hit, they do not fall back, but they pierce through a shield, then cut through a heavy iron corselet and wing their way through and out at the other side. So violent and ineluctable is the discharge of arrows of this kind. Such an arrow has been known to pierce a bronze statue, and if it hits the wall of a very large town, the point of the arrow either protrudes on the inner side or it buries itself in the middle of the wall and is lost. Such then is this monster of a crossbow, and verily a devilish invention. And the wretched man who is struck by it, dies without feeling anything, not even feeling the blow, however strong it be. — Anna Komnene

So it probably looked like the following, of course the trigger lever wouldn't be THAT long whereas the powerstroke would be longer, but you get the point:

I also think that based on the pinching method shown in Timurid, Safavid and Mughal crossbow, that it may be possible to create a very simple trigger based on pushing by the thumb.

This method allows a long powerstroke but cannot have a strong draw weight because it's a single axle lever and the lever's short.

"The "trigger pull" describes the force necessary to overcome this resistance-to "pull" the trigger (in fact, a trigger should be squeezed, not pulled). The more strength required to squeeze the trigger, the greater the risk of shaking or jerking the weapon. The introduction of one or more levers between the trigger and the part that releases the bowstring allowed the strength of the bow to be distributed along several axes restraining the internal levers, resulting in a notably reduced trigger pull. This development, begun in the decades before 1500, produced the two-axis lock and, in the second half of the sixteenth century, culminated in the internal-release mechanisms ultimately involving four axes. Readying a crossbow with a four-axis lock and set trigger was a complicated process involving a number of time-consuming steps and an additional instrument, the pricker, or cocking pin."Even in the hands of an expert, the entire sequence could take several minutes........while the reduction of the trigger pull and the resulting sensitivity of the release mechanism dispensed with the need for long lever triggers, they also made accidental discharges more likely. The former lever trigger, or a metal fixture of the same shape, was now secured to the tiller in a fixed position to form a trigger guard. - A Deadly Art, European Crossbows, pg 32-33

From the same book:

Top one is the Medieval trigger, middle one is the two-axis lock trigger, bottom one is the four-axis lock trigger. Notice the trigger got shorter on the bottom one.
Last edited:
Aug 2019
Hi, have you got any info about estimated battlefield rate of fire of Han/Song dynasty crossbows? I know that It depended on many factors (for ex. skill and strength of shooter), but I mean only very rough estimation. I expect that countermarch slowed reloading down but probably with such heavy draw weight longer intervals were necessary.
Feb 2011
A Medieval hand drawn crossbow could shoot at 6 bolts per minute with moderate pacing, and a Medieval windlass crossbow could shoot at around 2 bolts per minute with the same pacing.
Based off of that, I would say that a Han hand drawn crossbow (no stirrups) could shoot at around 5 bolts per minute, a Song dynasty hand dawn crossbow could shoot at around 6 bolts per minute (because it has a stirrup), and a Han/Song leg drawn crossbow could shoot at around 4 bolts per minute. This is all assuming moderate pacing.

There is also the windlass crossbow which first appeared during the Han dynasty, which probably shoots at 2 bolts per minute. I only found one contemporary pictorial of this type of crossbow, so such types would have been few and far between.

Also I'm basing my estimations off of the following:
1. For hand/leg drawn crossbows, the time it takes to draw the string from rest to full-draw takes up a very very tiny portion of the entire reloading process
2. For windlass crossbows, the time it takes to draw the string from rest to full-draw, combined with the time it takes to unwind the windlass, takes up a very large portion of the reloading process
3. Because of point 1 powerstroke shouldn't matter that much in terms of reloading speed for hand drawn crossbows, but because of point 2 powerstroke significantly affects reloading speed of windlass crossbows
4. The Han/Song crossbow re-locks its trigger nut by pulling on the aiming lever. The Medieval European trigger nut is re-adjusted by physically rolling the trigger nut with the thumb until it locks back into place.
5. Han windlass crossbows used a single pulley system whereas at least some late Medieval windlass crossbows used a compound pulley. Han windlass crossbows also probably had longer powerstroke.
6. Single pulleys pull less weight (in this case draw weight) than compound pulleys, but given the same pull distance a single pulley can operate twice as fast as a double compound pulley and thrice as fast as a triple compound pulley.
7. In terms of reloading speed points 5-6 evens themselves out for Han/Medieval crossbows
Last edited:
Feb 2011
The below is a translation from the diary of a Song soldier named Zhao Wannian in 1207 AD. The diary is about the Jin siege of Xiangyang. The diary is named "A Record of the Defense of Xiangyang's City Wall". The translator is Julie Jane Avery: https://pdfs.semanticscholar.org/9ae4/e034363aebb6501bf528c4e9f465275d662a.pdf
Through investigation, we heard that the enemy feared the crossbow arrows the most. If an arrow struck, it [could] pierce a horse’s belly and pass through heavy armor. Our spies also learned that the enemy was short of [longbow] arrows, so that each time they came and attacked the city wall, they would secretly collect the [longbow] arrows on the ground that had been shot from on top of the wall, and reuse them to shoot into the city. The Honorable [Zhao] passed down an order that we must not [carelessly] release [even] one [longbow] arrow because he feared that they would be used by the enemy. Because of this he sought out and asked men, one by one, who were among the [ranks of the] [long]-bowman (and also the spearman, shield bearers, and swordsman) for those willing to change to being crossbowman. All of those who heeded [this call] were paid three strings of cash. Consequently, we obtained more than three-thousand men who changed to become crossbowman, and added them to the top of the wall.
[29Ae] The [siege and] defense of the city started in winter and ended in spring. The force of the crossbows gradually decreased, so we were afraid that [the arrows] would not be able to reach [targets] far away. Consequently we set up the practice of [attaching] a longbow to the back of the crossbow to help make up for the deficiency.149 Thus their force increased and did not decrease, and [the crossbows] were able to reach distant [targets]………………………
[29Bi] Each time that we engaged in battle, we would use no less than one-hundred thousand crossbow arrows in one day. Although there were [supplies of] crossbow arrows in the city, we still feared running short [of them]. Consequently, [Zhao] commanded us to seize the enemy’s arrows. We cut them to make crossbow arrows,152 and only lacked the plume feathers. Consequently, two inches below the thick end of the [arrow] we drilled one hole and stuck hemp through it to substitute for the plume. [The arrows] were both able to travel a long way, and especially able to penetrate objects.


I bolded the most interesting part. As stated in the beginning of this thread (post 1 and 5), crossbow draw weights are at its highest during dry winters, and lowest during hot humid summers because that's when the glue of the composite prod starts breaking down. The diary mentioned that as winter and spring ended, the crossbows decreased in power, so they offset the power loss by increasing the draw weight. This was done by adding a handheld bow to the back of the crossbow, which made it similar to a miniature two-prod ballista:

Cui Wei Xian Sheng Bei Zheng Lu :
During the hot summer, when it’s raining and there’s steam, the horn [composite prod] should be easily taken off, and use a wooden [simple prod] crossbow. During the spring and winter, where there is wind and bitter cold, the wood (simple prod) is heavy and sluggish, and horn [composite prod] crossbows must be used.

Andreas Bichler, who makes replica Medieval crossbows, also notes that his crossbow loses stiffness (and hence draw weight) during the hot humid summers, but gains the draw weight back during the dry cold winter.
Last edited:
Likes: Ichon

Similar History Discussions