Design and Rate Control of Large Titanium Alloy Springs for Aerospace Applications

Abstract

During the separation between satellite and launch vehicles, large steel springs are often used as compression separation spring sets in a catapult separation system. Replacing the steel springs with titanium alloy springs could reduce weight by about 50%. Although titanium alloy springs have been widely used in the aerospace field due to their excellent performance, there are few reports on the design of high-precision titanium alloy springs. The current spring design standards mainly focus on steel springs with helix angles between 5° and 9°, which are not applicable to titanium springs. Moreover, the change in spring rate with ambient temperature should also be considered. In this paper, β-C titanium alloy was used to design and prepare large compression separation springs, replacing steel springs in the catapult separation system. The design of titanium alloy springs took into account the big helix angle. The relationship between helix angle and the number of active coils was calculated. The parameters of titanium alloy springs were determined by the shear stress of the spring at working length. The effects of aging temperature and aging duration on the mechanical properties and modulus of β-C alloy were studied. By adjusting the aging process, the β-C alloy spring rate was controlled to meet the design requirements. The effect of ambient temperature on the mechanical properties and modulus of β-C titanium alloy were also investigated. It was found that as the ambient temperature increased, the rate of the β-C alloy spring gradually decreased.

1. Introduction

Since the invention of the first metastable β titanium alloy B120VCA (Ti-13V-11Cr-3Al) in 1956, its ultra-high tension strength (1724 MPa), low modulus (Elastic modulus = 103.4 GPa, shear modulus = 41.4 GPa), and excellent fatigue strength have made it the best candidate material for aviation springs [1,2,3,4,5]. Substituting titanium alloy springs for conventional steel springs in the aerospace field can provide many advantages in weight, size, and high corrosion resistance. Firstly, the density of titanium is about 60% of steel, and the elastic modulus and shear modulus of titanium alloys are about half of steel. Therefore, the number of active coils required for titanium spring is only half of steel springs. The weight reduction can reach about 70%, theoretically [3,4]. Secondly, the lower shear modulus of titanium alloy makes it possible to achieve the same load capacity in a shorter spring design. Titanium springs can theoretically save about 50% of space in the height direction [3], making them suitable for component design in narrow spaces. Moreover, titanium alloys have high corrosion resistance. Complex anti-corrosion treatment processes are not required for titanium springs. Even in harsh environments, titanium springs can ensure long-term stable service and reduce maintenance costs. In addition, titanium springs also have good damping characteristics, requiring less energy for acceleration and deceleration during movement. Lower energy can cause elastic deformation, making the movement smoother and more controllable. Therefore, titanium alloy springs are not only widely used in the aviation industry, but also in high-speed trains, racing cars, and snow motorcycles. Weight reduction, corrosion resistance, and the excellent damping characteristics of titanium alloy springs have been fully utilized.

The first commercial application of titanium alloy springs in the aviation industry began in 1970. McDonnell Douglas Company used Ti-13V-11Cr-3Al alloy springs for a DC-10 wide-body plane, which were mainly applied in the nose and main landing gear down-lock springs, elevator, and aileron control springs [1]. In 1971, RMI invented β-C (Ti-8V-6Cr-4Mo-4Zr-3Al) metastable β titanium alloy. In addition to the high strength and low modulus properties, β-C also had excellent cold working ability, making it easy to prepare springs with wires or small bars. β-C alloy gradually replaced Ti-13V-11Cr-3Al alloy, which had more processing difficulties. The initial use of β-C alloy springs was on Boeing 757 narrow-body planes as passenger door counterbalance springs, with a weight reduction of up to 66.6% [4]. Later, β-C alloy springs were also used in Boeing 777, A330, A340, and other aircraft. TIMETAL LCB is a low-cost β titanium alloy designed specifically for automotive coil spring applications. The LCB alloy springs were first used as a suspension spring in the Volkswagen Lupo FSI sedan in 2000, making the car more comfortable and fuel-efficient. Similar suspension springs have since been used in Ferrari Challenge Stradale sports cars and other vehicles [5,6].

During the separation between satellite and launch vehicles, large separation springs are used to eject the satellite from the launch vehicle [7,8,9]. The ejection velocity depends on the number of springs, their rates, and pre-compression. Large steel springs are often used as compression separation spring sets in a catapult separation system. Replacing steel springs with high-strength titanium alloy springs in a catapult separation system can achieve weight and space savings and can omit the surface anti-corrosion treatment. There are some important considerations for the design and preparation of large titanium alloy springs applied in the field of space industry. Firstly, the current standards, such as GB/T 23935-2009 “Design of cylindrical helical springs” [10], EN 13906-1 “Cylindrical helical springs made from round wire and bar—Calculation and design” [11], and IS-7907-1 “Design and calculation for springs made from circular section wire and bar” [12], mainly apply to steel springs, with a spring helix angle range of 5–9°. But, the helix angles often exceed 9° for titanium alloy springs because of the low elastic and shear moduli, and fewer coils. These standards may not meet the high-precision design requirements of titanium alloy springs, and the big helix angle should be taken into account during the design of titanium alloy springs. In addition, titanium alloy has a large deformation rebound [13]. The size variation after cold work and heat treatment makes it difficult to precisely control the rate of titanium spring. So, how to control the spring rate by heat treatment should be investigated. At last, the change in spring rate with ambient temperature should also be considered [14].

So far, there is no systematic research on the design and rate control of titanium alloy springs, which are of great significance for the safety and reliability of satellite separation from the rocket. In this paper, large β-C alloy springs were designed and prepared to replace the original steel springs in a catapult separation system. The control of spring rate was investigated by adjusting the aging temperature and aging duration. The influence of ambient temperature on the performance of titanium alloy springs was also studied.

2. Design and Preparation of Titanium Alloy Spring

2.1. Parameters of Original Steel Spring

Figure 1 shows a schematic diagram of original steel spring for spring sets in a catapult separation system. It is a right-handed cylindrical helical compression spring. The weight of a single spring is about 5.1 kg. The wire diameter (d) of the steel spring is 12 mm. The free length (L₀) is 616 mm, and the mean diameter of the spring (D) is 66 mm. The number of active coils (n) of the spring is 26. The number of total coils (n_t) is 28. The spring rate (R) is 28.0–28.8 N/mm. The spring needs to undergo a stress relaxation test of 48 h at working length (341 mm). Other technical conditions shall be executed in accordance with the standard QJ 1787-89 “Technical Conditions for Cylindrical Compression Springs” [15].

2.2. Design of Titanium Spring

The characteristic line of large-sized titanium alloy cylindrical helical compression springs can be approximated as a straight line, and the spring rate is considered to be a constant. According to the standards GB/T 23935-2009 [10] and EN 13906-1 [11] and IS-7907-1 [12], the equations for spring rate and shear stress are as follows:

$$R=\frac{F}{f}=\frac{G{d}^4}{8D^{3}n}$$

(1)

$$\tau =\frac{8KDF}{\pi d^3}=\frac{KGdf}{\pi D^{2}n}$$

(2)

where R is the spring rate; F is the load; f is the compression distance; G is the shear modulus of the spring material; d is the wire diameter of spring; D is the mean diameter of spring; n is the number of active coils; τ is the shear stress; and K is the spring curvature coefficient.

Actually, Equation (1) is a simplified version of spring rate, which is suitable for the steel springs with a helix angle range of 5–9°. While the helix angle is often greater than 9° for titanium alloy springs, the spring rate equation needs to consider the influence of the helix angle (α). The spring rate equation should be as follows [16]:

$$R=\frac{F}{f}=\frac{G{d}^4}{8D^{3}n}\times \frac{E\cos \alpha }{E{\cos }^2\alpha +2G{\sin }^2\alpha }$$

(3)

New variables of elastic modulus (E) and helix angle (α) have been added to Equation (3) for titanium alloy spring rate calculation. α is related to n, L₀, and the outside diameter D_o of the spring.

Before designing the titanium alloy spring to replace the original steel spring, two things need to be considered. One is that the rate of titanium spring (R) should be the same as the original steel spring. The other is that the size of the spring, including the outside diameter D_o and free length L₀, needs to be consistent to meet the installation conditions. It is worth noting that if E and G are treated as constant, all unknown variables in Equation (3) are functions of n and d.

β-C alloy springs have been widely used in the aviation industry. Their stability and reliability have been verified [1,2,3,4,5]. Therefore, β-C alloy was selected as the titanium spring to replace the steel spring in Figure 1. The basis of the β-C alloy used are as follows: E = 104 GPa and G = 40 GPa. The maximum value of τ = 840 MPa was used in reference [3] but, in this paper, the maximum value of τ was used as 800 MPa for the safety considerations. The spring rate was taken as the middle value R = 28.4 N/mm. The mean diameter of spring D = 78-d mm. According to Equation (3), the calculation results are shown in Figure 2.

As shown in Figure 2, with the decrease in the number of active coils, α increases gradually. When n ≤ 18, α exceeds 9° and the difference between Equation (1) and Equation (3) cannot be ignored for high-precision springs. The wire diameter decreases with the number of active coils. Fewer coils and a smaller wire diameter are both beneficial for weight reduction. However, the selection of n requires consideration of the maximum shear stress at working length. So, the torsional yield strength of the β-C alloy should be considered to select the appropriate parameters of the titanium spring.

Figure 3 shows the shear stress of the spring at working lengths of 366 mm and 341 mm, as well as the solid length (the length of helical springs when compressed until all the coils touch). As the working length decreases, the deformation of the spring increases, and the shear stress of the spring also increases. Based on the maximum working length of 341 mm, when the active number of coils is 18, the shear stress is 780 MPa, which is closest to the set value. The parameters for titanium alloy springs should be selected with an active number of coils n = 18, a total number of coils n_t = 20, and a spring wire diameter d of 13.0 mm. According to the selected spring parameters, the spring weight is calculated to be about 2.6 kg, which reduces the weight by more than 49% compared to the steel spring.

In addition, the fatigue performance and stress relaxation of titanium alloy springs need to be considered for titanium alloy springs used in aviation industry. Boyer R.R. et al. [1] reported that higher tensile strength often corresponds to higher torsional yield strength in the range of 1241–1620 MPa, but lower tensile strength has better fatigue performance. Hu J. et al. [14] reported a stress relaxation mechanism and life evaluation under long-term thermo-mechanical coupling conditions. The result indicated that the primary failure mode for helical compression springs under prolonged static thermo-mechanical coupling conditions is stress relaxation rather than fatigue failure. The large titanium alloy spring for compression separation spring sets in the catapult separation system was not recovered. So, a stress relaxation experiment was conducted, and the fatigue performance of β-C titanium alloy spring is not discussed in this paper.

2.3. Preparation of Titanium Springs

The β-C titanium alloy ingot was prepared by the Northwest Institute for Nonferrous Metals Research through three rounds of vacuum consumable arc melting. The ingot was forged through multiple rounds at different temperatures and was rolled for three rounds. The final pass of hot-rolling for the β-C alloy bars was conducted at 920 °C, and it was water quenched. Then, the bars were precisely machined to the target size by a centerless lathe to obtain bars with uniform diameters. The chemical composition of the β-C alloy bar is shown in

Table 1. Chemical composition of β-C alloy bar (wt%).

Al	V	Cr	Zr	Mo	Fe	O	N	H
3.54	8.18	6.08	4.13	4.01	0.05	0.097	0.01	0.0011

.

As shown in Figure 4, the cross-section microstructure of the β-C alloy bar is a typical equiaxed microstructure of metastable β titanium alloy. The equiaxed β grains are uniform and the average grain size is about 30 μm.

Two kinds of β-C springs were cold coiled by using a CNC spring coiling machine. The springback of titanium alloy springs during cold coil processing was relatively large. It was necessary to choose a suitable core rod and consider the amount of springback after unloading. In addition, the free length and outside diameter of the springs increased after heat treatment. Therefore, it was necessary to appropriately compensate for the total number of coils of the spring during cold coil processing. Both ends were tightened followed by flattening of the springs using the grinding machine by 3/4 coils. After aging treatment, these springs were finally shot peened to remove the surface oxide layer.

As shown in Figure 5a, β-C alloy springs (1#) with a total of 20 coils (n_t = 20) were made to replace the steel springs. In addition, a spring (2#) with the same wire diameter and outside diameter as 1# was prepared. The spring (2#) had a total of 6 coils, which was created for use in a high-temperature compression test. The verticality of springs 1# and 2# were well controlled, and the spring pitch was uniform and consistent.

3. Rate Control of Titanium Spring by Heat Treatment

Wire diameter is the most critical parameter affecting the spring rate. If the deviation of wire diameter is 1%, the load will produce a deviation of about 4%. For β-C alloy spring (1#), even if the deviation of the wire diameter is 0.1 mm, the change in spring rate is about 0.9 N/mm. For high-precision springs, the uniformity and consistency of the spring wire (bar) are very important. However, deviations often occur during the preparation process, resulting in the spring rate exceeding design requirements.

The elastic modulus E and shear modulus G of metal materials are related to the heat treatment process. In the spring design standard GB/T 23935-2009 “Design of cylindrical helical spring” [10], the shear modulus of steel wire 60Si2MnA and 50CrVA are both 78.5 GPa. However, experimental studies have shown that the heat treatment temperature has a significant effect on the shear modulus. When the aging temperature of 60Si2MnA is 450 °C, the shear modulus is 83.16 GPa, and when the aging temperature of 50CrVA is 600 °C, the shear modulus is 86.6 GPa, increasing by 5.94% and 10.32%, respectively. For β-C alloy bars, both aging temperature and aging duration can affect the elastic modulus and shear modulus. Therefore, the rate of titanium alloy springs can be controlled by adjusting the aging temperature or aging duration.

3.1. Influence of Heat Treatment Process on Mechanical Properties

The recommended range for the aging temperature of β-C alloy bar and wire in the standard AMS4957E is 482–566 °C, and the range of tensile strength is 1241–1379 MPa (for 9.52–15.88 mm diameter) [17]. The aging processes used in this paper are shown in

Table 2. Aging process of β-C alloy bars.

No	Temperature	Duration	No	Temperature	Duration
1	440 °C	10 h	8	480 °C	0
2	460 °C	10 h	9	480 °C	4 h
3	480 °C	10 h	10	480 °C	8 h
4	500 °C	10 h	11	480 °C	12 h
5	520 °C	10 h	12	480 °C	16 h
6	540 °C	10 h	13	480 °C	20 h
7	560 °C	10 h	14	480 °C	30 h

. The aging temperature ranges from 440 to 560 °C. The samples have a maximum aging duration of 30 h. The testing method for mechanical properties of β-C alloy bars and the processing of specimens were in accordance with GB/T 228 “Metallic materials—Tensile testing” [18]. Two samples with the same aging processes were used to complete the test, and the data used below are the average of the test results of the two samples.

Figure 6 shows the effect of aging temperature and aging duration on the tensile strength and elongation of β-C alloy bars. The mechanical properties of the original bar (solid-solution treatment state, with aging duration of 0 h) are shown in Figure 6b. The tensile strength is 907 MPa, and elongation is 20%. When the aging temperature is 440 °C, the strength is still less than 1241 MPa, even if the bar has been aged for 10 h. When the aging temperature reaches to 480 °C, the tensile strength of β-C bar is 1385 MPa, reaching its maximum value under the aging duration of 10 h. The tensile strength of the β-C bar gradually decreases under higher aging temperature. When the temperature reaches 540 °C, the tensile strength drops below 1241 MPa. The elongation of the β-C bar gradually increases with the increase in aging temperature, from 10% to 18%. As shown in Figure 6b, the tensile strength increases rapidly with aging duration in the initial 8 h. Afterwards, the tensile strength slowly increases and gradually stabilizes. The elongation of β-C bars rapidly decreases with aging duration. When the aging duration exceeds 8 h, the elongation slowly decreases and gradually stabilizes. Wang J. et al. [19] reported similar results on the effect of aging temperature and duration on the tensile strength of β-C bars. The tensile strength and yield strength of β-C bars first increase and then decrease with the increase in aging temperature, reaching their maximum values at 460 °C. The temperatures corresponding to the maximum strength reported in reference [19] and those in this paper are very close, and the difference of 20 °C may be due to different rolling temperature and aging duration.

3.2. The Influence of Heat Treatment on the Moduli of β-C Alloy

The rate calculation of titanium springs not only needs the shear modulus to be known, but also the elastic modulus [20]. The elastic modulus of β-C alloy bars can be determined through tensile tests. Figure 7 shows the effect of aging temperature and aging duration on the elastic modulus of β-C alloy bars. As the aging temperature increases, the elastic modulus gradually increases from 95 GPa (aging at 440 °C) to 105.5 GPa (aging at 480 °C). When the aging temperature is above 480 °C, the elastic modulus slowly decreases to 102.5 GPa. As the aging duration increases, the elastic modulus rapidly increases within the initial 8 h and then increases slowly. After 30 h of aging, the elastic modulus is 108.5 GPa.

The shear modulus (G) and elastic modulus (E) can be converted using Equation (4):

$$G=\frac{E}{2\left(1+\mu \right)}$$

(4)

where μ is the Poisson’s ration. The shear modulus value can be derived through Equation (4). For springs with optimal lightweight design requirements, it is better to carry out torsion tests in accordance with GB/T 10128-2007 “Method for Torsion Testing of Metallic Materials at Room Temperature” to determine the shear modulus and torsional yield strength of the material [21].

The elastic modulus of metal is not only related to its chemical composition, but also to its phase composition and microstructure [22,23,24,25,26]. Figure 8 displays the XRD patterns of β-C alloy samples with different aging temperatures and aging durations. The 8# sample corresponds to the aging process of No. 8 in Table 2, which is the solid solution treatment (ST) state of the β-C alloy bar. The XRD pattern of 8# shows only the sharp diffraction peaks of the β phase. Figure 4 shows the microstructure of 8#. Large-angle grain boundaries of the β phase are clearly corroded, but part of the grain boundaries is not corroded due to their small orientation difference.

After 10 h of aging treatment, the intensity of the diffraction peaks of the β phase decreases and widens, as shown in Figure 8a. New diffraction peaks of α phase are observed in samples 2#–5#, which means the α phase precipitates from the β phase. Especially at 520 °C, all the characteristic diffraction peaks of the α phase can be clearly observed, and the peak intensity ratios are consistent with the PDF card (No:42-1294#). That indicates that the precipitation process of the α phase has a random orientation. Moreover, as the aging temperature increases, the peak intensity of the α phase increases gradually. That means an increase in the proportion of α phase in the β-C alloy samples, which will lead to a variation in the elastic modulus [20].

Figure 8b displays the XRD patterns of the β-C alloy samples with their aging duration at 480 °C. After aging for 4 h, the intensity of diffraction peaks of the β phase decreases and widens, and weak diffraction peaks of the α phase are observed. The diffraction peaks’ intensity in the α phase increase with the aging duration. But, when the aging duration exceeds 12 h, the precipitation of the α phase is close to saturation, and the intensity of the α phase peaks remains consistent.

Figure 9 shows the microstructures of the β-C alloy samples with different aging temperatures and aging durations. When aging at 460 °C, the α phase precipitates within thermodynamic-unstable grains. A small amount of the α phase particles appears within the grains, as shown in Figure 9a. However, due to insufficient aging temperature, there are lots of bright α phase precipitate-free zones within the β grains. With the increase in aging temperature, the α phase could fully precipitate within 10 h, and the proportion of α phase precipitate-free zones decreased. The increase in volume fraction of the α phase with increasing aging temperature is consistent with the XRD data. Moreover, the morphology of the precipitated α phase transits from spots to lines, finally forming a network. The size of the precipitated α phase becomes coarser with the increase in aging temperature. When the aging temperature is low, the nucleation rate is high and it cannot grow easily due to slow diffusion. As the temperature increases, the nucleation rate decreases, but it can grow easily because of the fast diffusion [27].

As shown in Figure 9d, clear β grain boundaries are observed in the microstructure image, with numerous α phase precipitates in the form of dots and lines in the β grains. But, the precipitation is not sufficient due to the short duration, resulting in a significant amount of bright α phase precipitate-free zones. As the aging duration gradually increases, the proportion of precipitated phases increases significantly. The α phase precipitate-free zones gradually decrease, leaving only a few bright spots near the β grain boundaries.

According to the principle of precipitation strengthening, the second-phase particles have a significant strengthening effect on tensile strength. The size and distribution of the second phase particles affect the strengthening effect. Greater amount and smaller particle radius of the precipitation could benefit the strengthening effect [19,28,29]. During the process of aging, the α phase preferentially precipitates at β grain boundaries, which can improve the strength and lead to a rapid decrease in plasticity. When the aging temperature is low, the precipitated phase has insufficient energy to grow. So, the strength gradually increases with the proportion of precipitates. When the aging temperature is high, the precipitated phase grows rapidly, the precipitation strengthening effect weakens, and the alloy is more prone to deformation. It is obvious that microstructure has a significant impact on the property of deformation resistance. Thereby, the elastic modulus of the β-C alloy and rate control of springs can be achieved through heat treatment.

3.3. Control of Spring Rate by Heat Treatment

Figure 10 shows the loading and unloading curves of the β-C alloy spring (1#) after aging at 540 °C/10 h. The spring rate can be calculated from its loading curve, R = 25.2 N/mm. The spring rate is smaller than the design value R = 28.4 N/mm. Moreover, when compression distance reached 225 mm, the spring underwent plastic deformation and could not return to the free length after unloading. That is because the maximum shear stress of the β-C alloy spring exceeded the torsional yield strength.

The rate of the β-C spring can be controlled through heat treatment. Decreasing the aging temperature to 480 °C and increasing the aging duration can effectively improve the shear modulus of the β-C alloy. Meanwhile, the torsional yield strength will also be improved, which could effectively prevent permanent deformation of the spring at working length [20,22,23,24,25,26].

Figure 11 shows the loading and unloading curves of the β-C spring after aging at 480 °C/12 h. When the spring was loaded to the maximum working length of 341 mm (corresponding to a compression distance of 275 mm), there was no permanent deformation of the spring, and it could return to the free length after unloading. The rate of the spring was R = 28.32 N/mm, which meets the design requirements of the titanium spring. The spring was loaded to its maximum working length at room temperature and maintained for 48 h. During this process, the change in spring load was within the range of 8N. There was no permanent deformation after unloading.

4. The Influence of Ambient Temperature on the Rate of Titanium Alloy Springs

Ambient temperature has a significant influence on the mechanical properties of titanium alloys [14,20]. The elastic modulus and shear modulus of titanium alloys change with increasing temperature. Therefore, the influence of ambient temperature on spring rate should be considered when designing titanium alloy springs. Rateick et al. [30] reported the variation in shear modulus of β-C alloy springs with temperature, and the results showed that the shear modulus gradually decreased during the cooling process from 220 °C. However, this is opposite to the trend in the shear modulus of Ti-6Al-4V as a function of temperature [14]. In order to figure out the effect of temperature on the modulus and spring rate, a temperature range from room temperature to 380 °C was used in this paper.

Table 3. Mechanical properties of β-C alloy bars at different temperatures.

Temperature	Rm/MPa	Rp0.2/MPa	A/%	Z/%	E/GPa
RT	1388	1287.5	11.5	27.5	106
100 °C	1307	1174	11.5	33	98.1
200 °C	1271	1130	10.5	38	94.3
300 °C	1247	1068	11.0	38	88.7
380 °C	1238	1051	12.5	44	85.6

shows the data on the mechanical properties of the β-C alloy bars at different temperatures. The tensile strength of the β-C alloy bars decreases from 1388 MPa to 1238 MPa with the increase in temperature. The elastic modulus decreased from 106 GPa at room temperature to 85.6 GPa at 380 °C.

Spring (2#) in Figure 5b was used for a high-temperature compression test. The free length of the 2# was 144 mm. The outside diameter was 78 mm. The total number of coils was 6, and the active number of coils was 4.

The high-temperature compression test of the spring (2#) was completed in accordance with HB 7571-1997 [31]. When spring 2# reached the test temperature, it was maintained for 15 min and then compression testing was performed.

Table 4. Spring rate of 2# at different temperature.

Temperature	RT	100 °C	200 °C	300 °C	350 °C	380 °C
Rate (N/mm)	156.8	154.5	150.5	145.0	138.6	134.2

shows the spring rate of 2#, calculated from the high-temperature compression curves. The rate of the β-C spring gradually decreases with increasing temperature. When the temperature exceeds 300 °C, the reduction in spring rate become greater. The spring rate of 2# decreased by 14.4% at 380 °C. Therefore, it is necessary to consider the influence of ambient temperature on spring performance in order to ensure safety and reliability.

5. Conclusions

(1)

The current spring design standards are not applicable to big helix angle springs. The helix angle of titanium alloy springs is often greater than 9°, so new variables of E and α should be added to spring rate calculation equation.

(2)

As the aging temperature increases, the elastic modulus gradually increases until 480 °C, then the elastic modulus slowly decreases. As the aging duration increases, the elastic modulus rapidly increases within the initial 8 h and then increases slowly. By adjusting the aging temperature and duration, the rate of titanium alloy springs can be controlled within a certain range. The rate of the β-C alloy compression separation spring can meet design requirements after aging at 480 °C/12 h.

(3)

The tensile strength and elastic modulus of β-C alloy bars decrease with an increase in temperature. The rate of the β-C spring gradually decreases with increasing temperature below 300 °C. When the temperature exceeds 300 °C, the spring rate drops rapidly.