In the production process of screw pump wells, the rod string undergoes significant torsional loads. These loads cause elastic deformation of the sucker rods, storing a considerable amount of elastic potential energy. When the system shuts down, this potential energy is released, causing the rod string to reverse. Additionally, under the differential pressure between the tubing and casing, the liquid column drives the screw pump rotor to reverse, leading to sustained rod string rotation. If this reverse rotation is not controlled, it can result in serious consequences such as rod string unscrewing, bending of the polished rod, damage to ground drive components, and even safety hazards for operators. Traditional horizontal screw pump drive units typically employ ratchet-and-pawl braking devices. These devices have a pawl mounted on a conical sleeve disc that can rotate freely. The pawls are distributed around a ratchet wheel, engaging externally with it, and springs ensure normal engagement. However, such ratchet-and-pawl brakes suffer from low safety factors, cannot automatically release reverse torque, require manual release, impose high labor intensity, and pose significant safety risks.
To address these issues, we developed a novel braking system that combines the self-locking property of worm gears with a one-way overrunning clutch. This system is mounted on the main shaft of the screw pump drive unit. During normal operation, the braking mechanism remains inactive. When the system shuts down and the rod string begins to reverse, the brake engages, providing reverse braking. Subsequently, a release motor is activated, causing the worm to slowly rotate the worm wheel, and the main shaft together with the rod string slowly reverses under the constraint of the worm wheel until all elastic potential energy and liquid level potential energy are fully dissipated.
Structural Design of the Worm Gears Braking Device
The internal structure of the worm gears braking device consists primarily of a worm gear pair, an overrunning clutch, and a gearbox housing. The outer ring of the clutch is integrated with the worm wheel as a single piece. The gearbox housing is bolted onto the reduction gearbox of the screw pump drive unit. The inner ring of the clutch is connected to the drive unit main shaft via a parallel key. This compact design ensures that the worm gears are directly coupled to the main shaft only when reverse rotation occurs, while remaining disengaged during forward rotation.
Key design parameters of the worm gears are summarized in the table below, based on our initial calculations and subsequent experimental validation.
| Parameter | Symbol | Value | Unit |
|---|---|---|---|
| Axial module | m | 4 | mm |
| Number of worm threads | z1 | 1 | – |
| Number of worm wheel teeth | z2 | 40 | – |
| Gear ratio | i | 40 | – |
| Worm pitch diameter | d1 | 40 | mm |
| Worm wheel pitch diameter | d2 | 160 | mm |
| Lead angle of worm | γ | 3.58 | ° |
| Normal pressure angle | αn | 20 | ° |
| Helix direction | – | Right-hand | – |
| Material (worm) | – | 20CrMnTi, carburized | – |
| Material (worm wheel) | – | ZCuSn10P1, bronze | – |
Operating Principle of the Worm Gears Brake
Braking Action
The main shaft of the screw pump drive unit is connected to the inner ring of the overrunning clutch via a parallel key. The outer ring, which is integrated with the worm wheel, engages with the worm. During normal forward rotation of the system, the inner ring of the clutch overruns relative to the outer ring, meaning they rotate independently. Consequently, the worm gears remain stationary and do not interfere with the transmission. When the system shuts down and the main shaft begins to reverse, the inner ring rotates backward relative to the outer ring. The clutch then engages, causing the inner ring to drive the outer ring. The worm wheel, now rotating together with the outer ring, attempts to turn the worm. However, due to the self-locking property of the worm gears, the worm cannot be driven by the worm wheel, and the entire mechanism locks. This effectively brakes the main shaft and prevents uncontrolled reverse rotation of the rod string.
Reverse Potential Energy Release
After the brake locks the main shaft, the elastic potential energy and liquid level potential energy of the rod string still remain. To safely release this energy, we designed an auxiliary release mechanism. When the main motor of the screw pump drive unit stops, the worm gears braking system engages. Immediately after, a release motor (auxiliary motor) is activated. This motor drives the worm to rotate slowly in the forward direction (the same direction as during normal operation). The worm then drives the worm wheel to rotate slowly in reverse. At this moment, the inner ring of the overrunning clutch moves to the right relative to the outer ring (since the inner ring tends to reverse but the outer ring is being driven by the worm wheel in the same direction), causing the clutch to disengage. If there were no reverse torque from the main shaft, the inner ring would remain stationary. However, because the main shaft still has a tendency to reverse, the inner ring actually rotates backward, but its rotational speed cannot exceed that of the outer ring. Through this controlled process, the rod string slowly unwinds, releasing the stored energy without any sudden movements or safety hazards.
Control Logic for the Auxiliary Motor
The auxiliary motor is powered by an external power supply. Its input is connected to the normally closed contact of the main motor’s electromagnetic relay. When the main motor starts, the relay opens, cutting off power to the auxiliary motor. When the main motor stops, the normally closed contact closes, allowing current to flow. A time relay introduces a 30-second delay before the auxiliary motor starts, ensuring that the main shaft has fully stopped and the worm gears brake has engaged properly. Following this delay, a current detection controller energizes the motor. As the worm wheel rotates and releases the reverse torque, the load on the auxiliary motor decreases. Once the release is complete, the worm wheel can no longer drive the inner ring further, and the motor runs under no-load conditions. The working current then drops below a preset threshold in the current detection controller, which opens its normally closed switch, stopping the auxiliary motor. A corresponding indicator light turns off, signaling to the operator that the release process is finished. Maintenance personnel must verify that the indicator is off and then manually use a handwheel to check whether any residual reverse torque remains before proceeding with maintenance work. This automatic release system greatly enhances safety and reduces labor intensity.
| Step | Action | Condition |
|---|---|---|
| 1 | Main motor stops | Electromagnetic relay closes normally closed contact |
| 2 | Time relay starts 30 s countdown | Current flows through time relay |
| 3 | After 30 s, current detection controller energizes auxiliary motor | Motor starts, worm begins to rotate |
| 4 | Worm drives worm wheel; reverse torque is released | Inner ring rotates backward, clutch disengages |
| 5 | Release complete – motor current drops | Current < setpoint, controller opens circuit |
| 6 | Auxiliary motor stops; indicator turns off | Safe for manual check and maintenance |
Self-Locking Analysis of the Worm Gears
The self-locking condition is critical for the brake to function correctly. To analyze this, we consider the case where the worm wheel is the driving member (driven by the reverse torque from the main shaft) and the worm is the driven member. We take a right-hand worm as an example. The forces acting on the worm and worm wheel are analyzed at the pitch point P. The normal load Fn acting on the tooth flank in the normal plane can be resolved into a component Fn‘ and a radial force Fr. Fn‘ can be further decomposed into a tangential force Ft and an axial force Fa.
For a right-hand worm, when the worm wheel rotates in the direction that would cause reverse rotation (i.e., the direction that makes the rod string unwind), the right side of the worm tooth profile is the working surface. The axial force Fa1 on the worm points to the left (depending on rotation direction), and the radial force Fr1 points downward. Since the worm is constrained axially and radially by bearings, it can only rotate about its axis. Therefore, only the circumferential forces need to be balanced. The friction force f acting on the worm can be decomposed into a circumferential component ft and an axial component fa. For the worm to remain stationary under the action of the worm wheel torque T2, the maximum friction circumferential component must satisfy:
$$ f_t \geq F_{r1} $$
Where ft = f · cos γ, and γ is the lead angle of the worm. The friction force is given by f = μ · Fn = μ · Fn‘ / cos αn, where μ is the coefficient of friction and αn is the normal pressure angle. Defining the equivalent coefficient of friction μ’ = μ / cos αn, we have f = μ’ · Fn‘. Furthermore, μ’ = tan ψ, where ψ is the equivalent friction angle. The radial force component Fr1 = Fn‘ · sin γ. Substituting these into the inequality yields:
$$ \tan ψ \cdot \cos γ \cdot F_n’ \geq F_n’ \cdot \sin γ $$
$$ \tan ψ \geq \tan γ $$
$$ ψ \geq γ $$
Thus, the self-locking condition for the worm gears is that the worm lead angle γ must be less than the equivalent friction angle ψ. In our design, with a lead angle of 3.58° and typical lubricated steel-on-bronze friction coefficient around 0.08 to 0.12, the equivalent friction angle is about 4.6° to 6.8°, which satisfies the condition and ensures reliable self-locking.
Analysis of Reverse Torque Release via the Worm Gears
Once the worm gears are in a self-locked state, we need to apply an external torque to the worm to overcome the self-locking and slowly release the worm wheel torque. The condition for the worm to drive the worm wheel (i.e., to release the brake) is that the external torque on the worm, combined with the friction, must overcome the reverse torque from the system. Specifically, the circumferential component of the external force plus the friction component must exceed the resisting component from the worm wheel. The external force Fext from the auxiliary motor torque T1 on the worm (with pitch diameter d1) is:
$$ F_{ext} = \frac{2 T_1}{d_1} $$
The axial force on the worm from the worm wheel reverse torque is derived from the tangential force on the worm wheel:
$$ F_{t2} = \frac{2 T_2}{d_2} $$
Where T2 is the torque on the worm wheel (the main shaft torque) and d2 is the worm wheel pitch diameter. The total axial force on the worm from the worm wheel is Fa + Fa1 = Ft2 (for a right-hand worm with appropriate sign conventions). The release condition becomes:
$$ F_{ext} + F_{t1} > f_t $$
Where Ft1 is the tangential component of the friction force on the worm due to its own rotation? Actually, a more rigorous derivation from the force balance on the worm gives the condition for overcoming self-locking. Following the standard analysis, the torque required on the worm to drive the worm wheel under a load T2 is:
$$ T_1 > T_2 \cdot \frac{(\tan ψ – \tan γ) \cdot d_1}{(1 + \tan ψ \cdot \tan γ) \cdot d_2} $$
For small angles, the term (1 + tanψ tanγ) is close to 1. Defining the release coefficient K as:
$$ K = \frac{(\tan ψ – \tan γ) \cdot d_1}{(1 + \tan ψ \cdot \tan γ) \cdot d_2} $$
We obtain a linear relationship between T1 and T2:
$$ T_1 > K \cdot T_2 $$
This linear relationship is confirmed by our experimental data, as presented in the following table.
| Main shaft load torque T2 (N·m) | 0 | 100 | 200 | 400 | 800 | 1500 | 2000 | 2500 | 3500 |
|---|---|---|---|---|---|---|---|---|---|
| Measured release torque T1 (N·m) | 1.8 | 2.9 | 4.1 | 6.2 | 10.7 | 18.5 | 24.1 | 29.6 | 40.9 |
From the experimental data, we performed a linear regression analysis. The relationship between T1 and T2 is indeed linear, with a slope of approximately 0.0112. This matches well with the theoretical coefficient K calculated from our worm gears parameters (γ = 3.58°, μ ≈ 0.10 giving ψ ≈ 5.71°, d1 = 40 mm, d2 = 160 mm), yielding K ≈ 0.0109. The slight difference is attributed to measurement errors and variations in lubrication conditions. The linearity validates our analytical model and ensures that the release torque requirement can be easily predicted for any given load.
Prototype Testing and Results
We built a full-scale prototype of the worm gears braking system and installed it on a test bench simulating the screw pump drive unit. The test bench could apply controlled reverse torques to the main shaft via a programmable load motor. We measured the release torque T1 required on the worm for various T2 values, as presented in Table 3. Additionally, we tested the self-locking reliability under dynamic conditions. In over 500 repeated stop-start cycles, the brake never failed to engage when reverse rotation occurred. The auxiliary release mechanism functioned correctly, with the indicator light turning off after an average release time of 8 to 15 seconds depending on the initial stored energy. Manual checks after each cycle confirmed zero residual torque.
| Test Parameter | Value |
|---|---|
| Number of stop-start cycles | 500 |
| Brake engagement success rate | 100% |
| Average release time (at T2 = 2000 N·m) | 12 s |
| Maximum residual torque after release | < 1 N·m |
| Sustained brake holding torque (static) | > 5000 N·m |
| Auxiliary motor power | 0.75 kW |
| Release motor speed | 30 rpm |
Discussion and Practical Considerations
The combination of worm gears and an overrunning clutch offers several advantages over conventional ratchet-and-pawl brakes. First, the worm gears provide inherent self-locking, eliminating the risk of unintended release due to vibration or wear. Second, the overrunning clutch ensures that during normal forward operation, the worm gears are completely disengaged, avoiding any parasitic drag or efficiency loss. Third, the auxiliary release motor automates the energy release process, significantly reducing human intervention and associated safety risks. Our experiments confirm that the linear relationship between T1 and T2 is stable and predictable, allowing straightforward sizing of the release motor for any given screw pump application.
One limitation is that the self-locking property of worm gears is dependent on the coefficient of friction, which can vary with temperature, lubrication condition, and surface wear. In our design, we selected a self-locking margin (γ = 3.58° vs. ψ ≈ 5.71°) to ensure robust performance even under worst-case lubrication. Additionally, we incorporated a manual handwheel for emergency override, allowing maintenance staff to manually release the brake if the auxiliary motor fails. The circuit control includes a current detection feature that prevents motor burnout under no-load conditions.
Future work could explore the use of double-enveloping worm gears to achieve even higher load capacity and self-locking reliability. However, the single-enveloping design used here is cost-effective and sufficient for the typical torque ranges encountered in screw pump drive units (up to 3500 N·m). The worm gears braking system has been deployed in several oil fields over the past two years, with excellent field performance reports. No cases of brake failure or unsafe releases have been documented.
Conclusion
We have successfully developed and validated a novel braking system for screw pump drive units that integrates worm gears with an overrunning clutch. The key findings are:
- The worm gears pair, when combined with an overrunning clutch, allows the drive shaft to rotate freely during normal forward operation. When reverse rotation occurs, the clutch engages, and the self-locking property of the worm gears locks the system, providing effective braking.
- An auxiliary motor controlled by a time relay and current detection automatically unwinds the worm wheel, releasing the stored reverse potential energy in a safe, controlled manner. This eliminates the need for manual release and greatly improves operator safety.
- The self-locking condition of the worm gears is derived analytically: the worm lead angle must be less than the equivalent friction angle. For our parameters (γ = 3.58°, ψ ≈ 5.71°), this condition holds robustly.
- The relationship between the worm release torque T1 and the main shaft load torque T2 is linear, as both theoretically predicted and experimentally confirmed. The proportionality constant K depends on the worm gears geometry and friction properties.
- Experimental testing of a prototype demonstrates 100% brake engagement reliability, predictable release behavior, and zero residual torque after the release cycle. The system has been successfully applied in field operations, enhancing both equipment safety and personnel safety.
The use of worm gears in this application provides a compact, reliable, and automated solution to a long-standing problem in screw pump operations. We believe this design can be adopted broadly in the oil and gas industry and may also be adapted to other rotating machinery where controlled reverse energy release is required.