The Design and Testing of a Special Drinker for Meat Ducks Based on Reverse Engineering

Abstract

Background: Intensive poultry production requires highly efficient drinking systems to ensure both animal welfare and production performance; however, conventional drinkers for meat ducks often suffer from design deficiencies that compromise drinking efficiency and result in significant water wastage. Objectives: To address the drinking water demands in intensive waterfowl farming systems, a specialized drinking device tailored for meat ducks was developed. Methods: The drinking habits of meat ducks were analyzed and the performance of the existing drinkers was evaluated. The deficiencies of the current drinkers were observed and identified by high-speed video, and the parameters of the head of the meat duck were obtained by reverse-engineering technology. Based on this analysis, a specialized drinker for meat ducks was designed, and its performance was confirmed through farming trials. Results: The static and dynamic flow rate tests showed that the output of the new drinker was consistent with the nipple drinker. When the valve rod was pushed upward, the new drinker did not output, which met the design requirements. The results indicated that, under a water pressure of 2.5 kPa, the water loss rate for the designed drinker was 27.4%, which was 15.3% lower than the loss rate of 42.7% observed with the traditional nipple drinker. Conclusion: This study develops a specialized drinker for meat ducks in intensive farming, by utilizing the biting drinking method and incorporating the three-dimensional characteristics of the heads of meat ducks, significantly increasing the effective drinking rate and reducing leakage during the drinking process.

1. Introduction

China is the largest producer and consumer of meat ducks in the world [1]. In 2023, the meat duck slaughter volume in the country approached 4.218 billion, accounting for over 80% of the global total [2]. The output value of these slaughtered ducks reached USD 20 billion, indicating the significant role of the meat duck industry within the livestock sector. With the continuous tightening of China’s environmental protection policies, traditional free-range water-surface farming has been prohibited, and meat duck farming has largely transitioned to a three-dimensional cage model [3]. This cage-based system significantly increases the farming density per unit area, which helps to optimize the use of space within shelters and facilitate the centralized processing of manure and waste [4]. In addition, it decreases labor requirements and enhances the intensive nature of farming, thus offering a competitive advantage in the large-scale development of the meat duck industry [5,6].

Currently, the cage farming model for meat ducks generally adopts poultry equipment designed for chicken farming. As a result, due to the incompatibility of the equipment, issues such as high feed-to-meat ratios and high culling rates have emerged [7]. The most prominent issues turn out to be insufficient water intake and water leakage from the drinker, which not only affects the feed-to-meat ratio [8] but also complicates manure management and causes other problems. Unlike chickens, ducks are waterfowl with a higher water demand. For ducks, drinking water is a crucial factor that influences their health and productivity [9]. The design of the drinker directly impacts drinking efficiency and health. Traditional drinker designs are often based on generic principles and fail to consider the specific biomechanical characteristics of ducks. This mismatch leads to low drinking efficiency, water wastage, and hygiene problems, all of which significantly impact the growth and development of meat ducks [10].

Currently, the nipple drinkers in use rely on the rigid sealing principle of fixed geometric shapes [11]. To enhance the drinking efficiency for meat ducks, numerous domestic and international companies and research institutions have worked on optimizing and innovating duck drinkers. Although many manufacturers have made various angle adjustments to optimize the meat duck drinker, the above-mentioned studies have not considered the specific morphology and drinking behavior of meat ducks. As a result, issues such as insufficient water intake, leakage during drinking, and excessive leakage leading to environmental pollution still occur during actual usage, which makes the widespread adoption of the drinker quite challenging [12,13].

In order to resolve the above-mentioned issues, this study conducted an in-depth analysis of the drinking habits of meat ducks and identified the problems in using nipple drinkers. Using reverse-engineering techniques, a novel and specially designed nipple drinker for meat ducks was proposed, tailored to their drinking behavior and biological features. The goal of this work is to improve drinking efficiency, reduce leakage, and meet the water consumption needs of meat ducks.

2. Observation and Analysis of the Drinking Habits of Meat Ducks

2.1. Materials and Methods

A total of 120 healthy SM3 Cherry Valley meat ducks, aged between 20 and 30 days, were selected and raised in a three-dimensional farming system. The dimensions of the cages were 1.35 m × 1.35 m × 0.65 m, with 10 meat ducks per cage. The temperature in the experimental house was set at 20 °C ± 5 °C, with the humidity maintained at 60% RH ± 10% RH. The environmental parameters were monitored using conventional digital temperature and humidity meters (range: −50 °C to +70 °C for temperature with ±1 °C accuracy; 10%RH to 99%RH for humidity with ±5%RH accuracy), which were mounted on each cage. All the measurements were recorded each day throughout the experimental period. A commercial nipple drinker was used as the water source for the meat ducks. All husbandry procedures in this paper were in strict compliance with commercial farming practices; only normal animal behavioral observations were made in this study; and no ducks were killed as a result of the experiment.

High-speed video was utilized to analyze the drinking behavior of meat ducks. A high-speed camera (FASTCAM Mini UX100, Photron, Tokyo, Japan) was used to record the drinking actions of the ducks, with a filming speed of 2000 fps and the resolution being set to 1280 × 1024 DPI. The camera was equipped with a cold light illuminator, which featured a 120 W adjustable brightness white LED, serving as the light source to enhance the imaging quality [14]. The LED illumination was maintained at 110 lux, equivalent to standard farm lighting conditions. Preliminary experiments confirmed that this illumination intensity did not significantly alter the drinking behavior of the ducks. To minimize any stress impacts on duck growth, the ducks were acclimated in the observation arena for 1 h prior to the subsequent 2 h behavioral recording session, totaling a 3 h experimental protocol. The camera position is shown in Figure 1, with the camera placed 100 cm away from the duck cage.

2.2. Analysis of High-Speed Video Results

The analysis of the high-speed video results of the drinking action (Figure 2) reveals several typical action patterns when meat ducks use the nipple drinker, including biting, pecking, and headshaking. During the biting process (Figure 2a), the meat duck repeatedly bites the lower valve rod from the front or side, using its beak to activate the water flow. At this point, the valve rod is completely held in the duck’s mouth, and the duck uses the inner edge of its lower beak to move the lower valve rod to release the water, which then flows along the lower jaw of the duck into the throat. During the pecking process (Figure 2b), the meat duck uses its upper beak to directly peck or push the lower valve rod upward to release the water. In this case, the valve rod is entirely outside the duck mouth, with some water flowing into the duck’s mouth, while more water flows onto the ground. During the headshaking process (Figure 2c), the meat duck, after biting the nipple drinker and not releasing it, shakes its head to increase the water flow through a larger motion. At this point, the valve rod is lifted by the duck beak. During the headshaking, most of the water is thrown out of the duck mouth and falls onto the ground.

The observations used to compare the frequency of different drinking actions revealed that pecking occurred most frequently, followed by biting, with headshaking being the least frequent, which typically occurs only when the meat duck begins drinking or has a high demand for drinking water. Additionally, based on the high-speed video analysis, we noted that, except for the continuous “head-shaking” action, each single “bite” or “peck” action generally lasted 0.2–0.5 s, and these actions were repeated approximately three–five times in one drinking bout. Meanwhile, a comparison of the effective water intake for each action showed that the biting action resulted in the largest water intake. During the forward biting, most of the water released can be consumed by the meat duck, with the loss of only a small amount. However, since the duck head does not remain stationary during drinking, it does not always maintain the forward biting action and may exhibit sideward biting. This process is similar to a combination of pecking and biting, where the valve rod moves both upward and sideways simultaneously. This motion causes the water to flow directly through the beak and fall onto the ground, as shown in Figure 3, with the circled area indicating where the water droplets fall. When the meat duck drinks using headshaking and pecking actions, the head shakes so quickly that the upper beak directly presses against the water outlet, causing the water to flow through the outer side of the upper beak and drip onto the manure belt.

Based on the high-speed video results, the causes of water dripping and leakage during the drinking process of meat ducks using the drinker can mainly be attributed to the following factors: (1) the beak of the meat duck has a flattened structure, which is different from that of chickens, causing the failure of the water flow obtained through the drinker to be directed into the mouth; (2) There is a structural flaw in the nipple drinker. When the meat duck raises the tip of its beak to touch the lower valve rod of the nipple drinker to obtain water, not all of the water successfully enters the mouth. Some of the water flows down the outer side of the upper beak and drips onto the manure belt. The observation results indicate that the main issue with the current commercial nipple drinker stems from a mismatch between the structural design and the characteristics of the beaks of meat ducks. As a water-loving poultry species, the daily water intake of meat duck needs cannot be met by the current effective water supply of the nipple drinker [15]. Therefore, design improvements should consider controlling the water flow through the drinking actions of meat ducks to ensure that the water is effectively directed into the mouth and to reduce water wastage and leakage.

3. Design Scheme of a Special Water Drinker for Meat Ducks

3.1. Design and Modeling of the Water Outlet Structure of the Drinker

To change the drinking behavior of meat ducks and increase the occurrence of biting actions, the water outlet structure of the drinker was designed. The internal structure of the new drinker is shown in Figure 4. The internal structure of the commonly used nipple drinker in farming relies on rigid sealing because of the water pressure between the steel ball and the housing. Continuing to use this method does not reduce the occurrence of actions such as pecking or biting. Therefore, the internal structure of the drinker was modified to use a spring as a reset component. The spring presses against the lower valve rod to create the seal, ensuring that the water release is triggered only by biting and headshaking, reducing the occurrence of pecking actions that lead to low water intake. The lower valve rod has a round-headed rivet structure, and the round head was selected to reduce the contact area at the sealing point, ensuring stability during sealing. The head of the lower valve rod trigger has a semicircular structure with a diameter of 6 mm and a thickness of 3 mm, with a slot of 4.5 mm in length and 1 mm in width at the top of the round head to ensure proper water flow. According to the study by Huo Juan et al. [16] on the biting force of birds and poultry, small-bodied birds with long beaks have a biting force of approximately 8–10 N. Therefore, a tower-type spring with a wire diameter of 0.6 mm, an outer diameter of 8 mm, and a length of 25 mm, providing a triggering force of 8 N, can meet the drinking needs of meat ducks up to 42 days of age.

3.2. Design and Modeling of the Drinker Shell

Meat ducks exhibit noticeable headshaking behavior during drinking. To reduce the occurrence of headshaking and prolong the biting time, a protrusion that conforms to the internal curvature of the lower beak was added to the biting area beneath the drinker shell. This ensures that the design of the drinker shell fits the duck beak, which enhances the comfort and acceptance of the biting action and reduces leakage.

Based on this design concept, biting was chosen as the primary drinking method for meat ducks. To accurately replicate the geometry of the lower beak of ducks and ensure that the drinker shell fits the duck beak while improving the comfort and acceptance of biting, reverse-engineering techniques were used to obtain the three-dimensional point cloud parameters of a duck head. The point cloud data can precisely describe the complex three-dimensional geometry of the duck beak.

3.2.1. Acquisition of the Point Cloud Data for the Meat Duck Beak

The experiment selected 30 mature meat duck heads at the age of 42 days and performed feather removal. A handheld laser 3D scanner (3DSHandy-Compass, Shuzao Technology, Shanghai, China) was used to scan the duck heads in three dimensions, with a scanning accuracy of 0.02 mm. The duck head samples were placed for horizontal cutting to obtain the sample of the lower beak of the meat duck, as shown in Figure 5. The scanned point cloud data and surface data were saved for further processing [17].

Using the reverse-engineering software Geomagic Studio 2013, the model was optimized and assembled, and the surface patches were constructed [18]. After fitting, a NURBS surface was formed, as shown in Figure 6.

After pre-processing the lower jaw surface of the meat duck, the extracted surface still did not conform to a fixed mathematical model visually. The surface fitting function in the MATLAB R2022a toolbox was used to fit the lower jaw surface of the meat duck. Polynomial fitting was applied to the feature surface point cloud. By adjusting the powers of x and y and observing the fitting coefficient R², the power degree with the highest fitting accuracy was selected to determine the equation of the fitted feature surface.

Based on the results in

Table 1. Fitting results of surfaces with different exponents as independent variables.

xnym	SSE	RMSE	R2
x1y1	4755	1.4111	0.73999
x1y2	4100	1.3109	0.77580
x1y3	4088	1.3095	0.77647
x1y4	4024	1.2998	0.77995
x2y1	2222	0.9651	0.87847
x2y2	1137	0.6903	0.93786
x2y3	1034	0.6589	0.94350
x2y4	1004	0.6496	0.94510
x3y1	2120	0.9429	0.88410
x3y2	1042	0.6614	0.94284
x3y3	1025	0.6560	0.94398
x3y4	932	0.6261	0.94905
x4y1	1894	0.8961	0.89641
x4y2	951	0.6323	0.94799
x4y3	933	0.6265	0.94898
x4y4	930	0.6257	0.94914

Note: xⁿ, y^m represent the maximum exponents of x and y in the fitting equation, where n is the maximum exponent of x and m is the maximum exponent of y; SSE represents the sum of squared errors; RMSE represents the root mean square error; and R² represents the coefficient of determination.

, when the exponent of the independent variable x is chosen as 4 and the exponent of y is chosen as 3, the SSE is 933, the RMSE is 0.6265, and the coefficient of determination (R²) is 0.94898. At this point, the fitting equation for the surface is as follows:

$$\begin{gathered} z_1=f(x,y)=-1.85x-2.749y+1.758x^2+1.678xy+0.6771y^2+0.1484x^3 \\ +0.2972x^{2}y+0.06902x{y}^2-0.1135y^3 \end{gathered}$$

(1)

The MATLAB fitted surface of the lower jaw surface of meat ducks is shown in Figure 7, where the values of x and y have a range of [−15, 10] and [0, 70], respectively. The dependent variable z varies within the range of [−20, 5] in this region. The standard deviation between the fitted surface model and the original model is 0.0154 mm, and the fitted surface mathematical model closely approximates the biological surface shape.

3.2.2. Modeling of the Drinker Shell

In order to prevent the headshaking action of meat ducks in the process of drinking water, the shell of the water drinker was improved. At the contact part between the lower beak and the occlusal part under the water drinker shell, the bulge that conforms to the inner surface of the lower beak was increased to prolong the occlusal time and reduce the occurrence of swinging action. Figure 8 shows the model of the water drinker shell. The overall length of the water drinker is 41.5 cm, and the shape of the outlet end is designed as a flat mouth type, with a width of 11.5 cm. According to the research on the three-dimensional structure of the head of meat ducks, the curved mathematical model of the lower beak of meat ducks was obtained, because the beak length of meat ducks at the brooding stage was about 15.0 cm. In order to ensure that the beak can hold the lower valve rod and the water outlet end during the drinking action, an arc beak surface can be added at the 12 mm of the rear of the water drinker shell, and the added surface can be connected with the shell by smoothing. The fitting equation of the lower surface is as follows:

$$\begin{gathered} z_1=f(x,y)=-1.85x-2.749y+1.758x^2+1.678xy+0.6771y^2+0.1484x^3 \\ +0.2972x^{2}y+0.06902x{y}^2-0.1135y^3 \end{gathered}$$

(2)

3.3. Overall Structure and Working Principle of the Drinker

The improved structure of the drinker is shown in Figure 9, which mainly consists of an end cap, lower valve rod, spring, and housing. In the resting state, the valve rod of the drinker is sealed under the combined action of water pressure and the spring. After the meat duck bites the lower valve rod, the valve rod tilts, creating a gap between the valve rod and the end cap, allowing water to flow through the gap. After the biting action ends, the lower valve rod blocks the water outlet of the end cap under the support of the spring when the meat duck releases its beak, which helps to stop the water flow and ensure the drinker is sealed again.

3.4. Manufacturing of Prototype for Meat Duck-Specific Drinker

The prototype of the drinker shell was manufactured using 3D printing technology, with the Li Chuang 3D L5-800 3D printer. White toughened resin was chosen as the processing material (Figure 10). The 3D-printed drinker sample had a width of 15 mm and a length of 50 mm. The internal components such as the upper and lower valve rods and steel balls were processed through subcontracted machining, using 06Cr19Ni10 stainless steel, which is corrosion-resistant as specified by GB/T 1220-2007 [19]. After assembly, the drinker exhibited a good sealing performance, allowing for the next step of performance testing.

4. Performance Testing of Water Drinker

4.1. Laboratory Test

After the drinker was assembled and its sealing performance was verified, it was installed onto the experimental water line for flow rate testing under both static and dynamic conditions (Figure 11). In a laboratory environment, repeated forces were applied to the drinker to simulate the actual behavioral actions of meat ducks during drinking, ensuring that the drinker could maintain a stable water output performance under different usage scenarios.

The water flow rate of the drinker in the farming facility is controlled by a pressure regulator at the end of the water line. A water level tube with a scale is placed above the pressure regulator, with different water levels corresponding to different pressure levels [20]. The test results showed that the static water flow rate of the new type of drinker ranged from 60 to 125 mL/min when the water level was between 100 and 700 mm (Figure 12).

In the static flow rate test, the water outlet pressure of the drinker was set to be 2.5 kPa. Under this condition, the lower valve rod trigger of the drinker was pressed to the open state. Two actions were performed: pressing the valve rod and pushing the valve rod upward. The test was maintained for 60 s, and the water output was recorded. Each test was repeated 10 times, and the average value was taken as the final result.

In the dynamic flow rate test, the biting action of the meat duck was simulated to test the responsiveness of the drinker and its stability under the biting action mode. The valve rod of the drinker was pressed from the side at a certain frequency to simulate the fast biting and drinking action of the meat duck. The test lasted for 60 s, with each press lasting 1 s and a 1 s interval between presses. The total water output during the test was recorded, and the average value was calculated after multiple measurements.

Figure 13a shows the results of the static flow rate test. The results indicate slight differences in water output under the two action modes. When the valve rod was pressed, the water output from the drinker was more stable, and the output was consistent between the two drinkers, with an output of approximately 82 mL, corresponding to a static flow rate of 82 mL/min. However, due to the special design of the new drinker that limits the upward movement of the valve rod, the water output was 0 mL when the valve rod was pushed vertically upwards. In contrast, the nipple drinker also allowed water to flow when pushed upward, with an output of approximately 100 mL.

The static water output standard for the nipple drinker is as follows: static water output = week age × 10 mL/min + 30 mL/min [21]. For meat ducks at the age of 5 weeks, the water output needs to be above 80 mL/min. The test results show that the static water output of the new drinker meets the water output standard for waterfowl drinkers, ensuring the optimal growth performance for meat ducks under these flow conditions.

Figure 13b shows the results of the dynamic flow rate test. After 60 s of continuous biting action, the average water output was 50.7 mL, with no jamming occurring. The average water output for the nipple drinker was 52.6 mL; it was statistically found that there was no significant difference in the average water output between the new and nipple-type drinkers under the biting action (p > 0.05).

4.2. Breeding Test

Since the newly hatched meat ducks were not accustomed to using nipple drinkers, a bell-shaped drinker was used for feeding, and the ducks were trained to use the drinker before 10 days of age [22]. A total of 200 healthy Cherry Valley meat ducks, aged 10 days, were randomly divided into two experimental groups, with 100 ducks in each group (sexes not considered). Group I used the conventional nipple drinker, while Group II used the new specialized drinker. During the experiment, all conditions, such as stocking density, temperature, and feeding method, were kept consistent across both groups, except for the drinking conditions. The experimental period lasted 30 days.

Both experimental groups used water tanks for the water supply, with the water levels marked on the tanks. The amount of water in the tanks was recorded daily, and the average daily water consumption per duck was calculated. Water leakage was measured once a day. In order to measure the amount of leakage when using the nipple drinker, a specialized leakage collection device was installed beneath the drinker (Figure 14). The main body of this device was a water collection tray measuring 150 mm × 900 mm × 200 mm, capable of holding approximately 13 L of water. Openings at the top and sides of the device allowed secure attachment to the cage using clamps. An independent drainage outlet was located at the upper right corner to facilitate the discharge and measurement of the excess water collected within the tray. To prevent the ducks from directly contacting or re-drinking the collected water, a protective railing structure was added around the tray. The specific measurement procedure was as follows: The tray was first positioned under the nipple drinker and fixed in the cage through the top openings. Subsequently, when the daily water consumption was recorded, the water that accumulated in the tray was drained through the upper right drainage outlet and weighed. By recording the collected water volume during the experiment, the leakage level of the nipple drinker could be accurately evaluated, thereby allowing the calculation of the actual water intake by the ducks.

The cages used in the experiment had the dimensions of 1000 × 800 × 300 (mm), with each cage equipped with an independent water tank of 10 L, which had water level and capacity scales for precise measurements. A water trough with a protective net was installed beneath each drinker to collect water droplets. Nine cages were set up per group, with 10 meat ducks in each cage and five drinkers per cage. Additionally, a dedicated reserve cage with dimensions of 2000 × 1100 × 300 (mm) was set up, equipped with 10 drinkers and housing 20 meat ducks. If any ducks died during the experiment, replacements were added from the reserve cage to maintain the number of ducks in the experimental group.

Result and Analysis

Figure 15 compares the effective drinking volume and total water consumption of the two groups of drinkers, where Group I represents the nipple drinker group and Group II represents the new type of drinker group. It can be observed that there is no significant difference in the total daily water consumption per duck between the two groups (Figure 15b), but the effective drinking volume of the new type of drinker group was significantly higher than that of the nipple drinker group after 20 days of age (Figure 15a). Throughout the entire farming period, the average effective drinking rate for the nipple drinker group (Group I) was 57.3%, while, for the new type of drinker group (Group II), the effective drinking rate increased to 72.6%.

Comparing the drinking data at different ages (

Table 2. Water consumption test results of different drinking water devices.

	Day	Effective Drinking Volume L	Water Yield a	Effective Drinking Water Rate η
1	10	157.4	223.5	70.4
	15	217.1	324.8	66.8
	20	288.4	465.6	61.9
	25	313.2	549.5	56.9
	30	328.2	663.5	49.4
	35	375.4	798.8	46.9
	40	423.2	865.1	48.9
Average				57.3
2	10	168.7	236.6	71.3
	15	245.4	349.8	70.1
	20	363.5	493.7	73.6
	25	446.3	583.6	76.4
	30	511.8	705.2	72.5
	35	547.1	774.1	70.6
	40	613.9	832.7	73.7
Average				72.6

), during the 30-day experimental period, the effective drinking rate for both drinkers exceeded 70% at the age of 10 days, as the meat duck beak had not fully developed at this stage, and the long beak had little effect on drinking. For Group I, the effective drinking rate of the drinker gradually decreased as the ducks aged, from 70.4% to 46.9%. For Group II, from the age of 10 to 40 days, the effective drinking rate of the drinker remained stable, with an average effective drinking rate of 72.6%, significantly higher than that of Group I. Video observations revealed that, as the meat ducks aged, the frequency of upward pecking movements during drinking increased in Group I, leading to increased water leakage (an average loss rate of 42.7%) and a reduced effective drinking rate. In contrast, the new type of drinker in Group II limited the upward pecking action, maintaining a high effective drinking rate throughout the process.

In this experiment, the meat ducks using the new type of drinker had a significantly higher effective drinking volume compared to those using the nipple drinker (Figure 15a), while their water consumption remained nearly the same (Figure 15b). Observing the drinking behavior, it was found that the water loss rate due to dripping was 42.7% in the nipple drinker group, while, in the new type of drinker group, the water loss was only 27.4%, which represented a reduction of 15.3% compared to the nipple drinker. Despite no significant difference in water output performance, the drinking behavior of the meat ducks using the new type of drinker showed changes. The frequency and duration of the biting action in the new type of drinker group were significantly increased compared to the nipple drinker group, which aligns with the expected performance.

5. Conclusions

This study addresses the issue of water consumption by meat ducks using nipple drinkers in intensive farming environments through behavioral analysis and engineering design, followed by the design of a specialized drinker for meat ducks. The performance of the newly designed drinker was tested through farming trials and showed a better performance compared to the commonly used nipple drinker.

• The drinking behavior of meat ducks was recorded, and it was found that the biting action is the most efficient drinking action when kept in cages. Therefore, biting was selected as the method for water release in the new type of meat duck drinkers, achieving optimal effective drinking without the need for training the ducks.
• Using reverse engineering, an original model of the head of meat ducks was created, and its features were repaired and reconstructed. The characteristic surface equation for the lower jaw of meat ducks was obtained, which could be applied to the subsequent engineering designs.
• The improved and optimized meat duck drinking device showed a 15.3% reduction in water loss compared to the nipple drinker and features a simple structure. It can meet the drinking needs of meat ducks during the farming process.