At the beginning of the establishment of space stations, “Zarya”, the first space cabin, was launched by Russia on November 20, 1998 [1]. Space stations act as a platform for human beings to observe the earth and study astronomy and bio-science [2]. The lifecycle of a space station ranges from 10 years to 15 years on average. During this period, random faults may happen on any type of device. In this case, on-orbit maintenance is a very effective solution. On-orbit maintainability is regarded as a very important tool to extend the lifecycle of space stations, ensuring the safety of astronauts and lowering the maintenance costs [3-5]. On-orbit maintenance activity of International Space Station is shown in Figure 1.

Maintenance tasks are not equal to maintenance operation. A maintenance task always starts from the detection of the fault until the function recovery and completion of the task [6]. Engineers should perform maintenance design, analysis, and verification to ensure the process before the vehicle is launched [7], while maintainability verification acts as the critical step during the entire process of maintenance. Figure 2 presents the maintenance workflow and design projects.

It can be broadly justified whether the quality and quantity targets of the maintenance work are compliant with relevant requirements [8]. Nowadays, several methods of maintainability verification have been studied and used, including analysis validation, simulation analysis, ground simulated testing, and on-orbit flight testing [9]. A proper method of maintainability verification depends on several factors, including the features of the spacecraft, the requirements of the maintenance, and the development of the project.

This paper embarks on the mission of the China manned space station. It proposes the by-step classification on-ground testing methodology after reviewing and comparing the existing maintenance validation methodologies. Meanwhile, it also establishes a zero-gravity simulation separated validation platform that accommodates the internal and external cabin ORU. In addition, it enhances the feasibility, availability, and economical efficiency of the work of maintainability verification.

Maintainability verification is a tool to verify whether the maintenance work can suffice the requirements of the key performance indicator. Verified objects are always referred to as the whole products with the designated feature and can be maintained on-orbit.

To ensure the safety and reliability of the spacecraft during its long working period on-orbit, the maintainability verification of a manned space station must resolve the following two major issues.

One issue is enormous verified devices. Given the occasional failure of electronic products, as well as the abrasion and stuck of institutional products, it is common to put the on-orbit maintenance functionality on the design of important device and function that include GNC (Guidance Navigation Control), energy, environmental, and thermal control, as well as information management and astronaut residence support. There are different impacts and maintenance timing even between the same types of ORUs. To the electronic device, the validation must be separated due to the different related downstream equipment. To the fluid pipeline equipment, the validation also has to be separated because of the layout. Different layouts mean different visualization and accessibility. Therefore, the quantity of ORUs for an independent-flight-spacecraft can be up to hundreds, which means there will be thousands of ORUs on the space station. Strictly, verification for each ORU cannot be substituted for other ORUs.

The second issue is the difficulty in testing and verification. It is a challenge to maintain the ORU including a single board of electronic products, fluid pipeline equipment, institutional products, equipment with precise installation, and large external cabin ORU. The maintenance should be verified thoroughly on the ground, which should take the requirements for the astronaut and manipulator into consideration as well as the complexity of the operations and environment restrictions [10]. It should be mentioned that verification on the ground is not equal to equipment assembly. The order of equipment assembly is prioritized by the comfortability of the operation. Meanwhile, for an aircraft in orbit, many of the key equipment cannot be shut down. It is more important to keep the station working than to make the operations convenient.

All figures should be numbered with Arabic numerals (1, 2, ⋯, n). All photographs, schemas, graphs, and diagrams are to be referred to as figures. Line drawings should be good quality scans or true electronic output. Low-quality scans are not acceptable. Figures must be embedded as pictures into the manuscript and not supplied separately. Lettering and symbols should be clearly defined either in the caption or in a legend provided as part of the figure. Figures should be placed at the top or bottom of a page wherever possible, as close as possible to the first reference to them in the paper. Large external cabin ORU maintainability needs two astronauts, and manipulator assistant is shown in Figure 3.

Accordingly, while designing the on the ground maintainability testing method and plan, factors including category, quantity, interface form, layout, and the challenges of on-orbit maintenance the designer should be considered to ensure reasonable project planning. To avoid fundamental changes in the product design and layout in the final project phase caused by maintenance plan adjustment, it is important to arrange the typical ORU verification at the very beginning, which usually is ORU of energy, environmental, and thermal control products as well as worse layout equipment.

In general, hierarchy of ORU can unfold in three levels, including product, subsystem, and system (aircraft). The validation of the product level relies on mechanical, electric, fluid, gas, and thermal interfaces, as well as the maintenance scheme of the product. The validation of the subsystem level relies on its working mode, maintenance tool, impact, and time consumption, as well as interfaces between this product and others, which belong to the same subsystem. The validation of the system level focuses on breakdown isolation, maintenance working mode setting, maintenance flow, system recovery, ORU transfer, maintenance impact of the whole aircraft, and maintenance tools arrangement.

Compared with the validation of the product level, verified items of the subsystem, even the system level, are obviously much more complex, involving breakdown isolation, separation of electricity and information, maintenance mode shift in-and-out of the aircraft platform, as well as maintenance operations, support, and time consumption. Therefore, verified items can be classified into some types. One type proves the ORU maintenance process, operation design and space, tools, and time consumption. The typical method is the flume experiment. The other type of verified items proves the default location, power supply and information separation design, maintenance working model, and scope of maintenance impact, and it usually combines with the integrative test.

The environment needed by on-orbit maintenance relies on a weightless environment and workplace.

The weightless environment is the critical factor to be considered during on-ground maintenance verification. Due to the power of the acting force and counter-acting force, astronauts must use a handle and stopper while performing maintenance work. It is shown in Figure 4 that there are many handles and stoppers in the cabin that are throughout the whole activity area.

The relative position between the astronaut and the apparatus, operation pit, and maintenance tools is different from the regular on-ground maintenance operation. In addition, because of the limited operation around the ORU, the coverage of the maintenance tool, standing position of astronaut, and operation scope are confined in a certain area. The serious limitations of activity space and visibility range are also applied to the maintenance program of external cabin ORU when the astronaut is in a spacesuit.

Therefore, when a maintenance verification plan is designed, it should simulate the scenario of human beings, apparatuses, and tools in a weightless environment to reflect the ORU layout environment, i.e., the astronaut visibility scope and activity area to ensure the testing result is the most similar as that in reality.

To stimulate the weightless environment of astronauts and ORU, on-ground testing methods have been extensively studied at home and abroad, among which such testing can be divided into mock tests and digital simulating tests. Digital simulating tests include computer picture simulation and VR simulation [11-12]. Mock tests include neutral buoyancy tank, parabolic flight, gas floating platform, hanging maneuver, drop tower, 1-g simulation, and hypobaric chamber [13-14]. This paper shows the comparison with the testing object, verified items, and the environment by some powerful methods.

As mentioned above, the quantity of ORUs for an independent-flight-spacecraft can be up to the hundreds. Almost all the ORUs should have their maintenance scheme under the cabin environment verified. If we use the maintenance verification methods in Table 1, long testing time consumption and excessive costs will not be avoided. Therefore, a proper on-ground maintenance verification method should be studied urgently to meet the requirements of the China space station.

Figures 5-7 show the different laboratory methods.

Regarding the analysis results of maintenance verification requirements, this paper proposes a by-step classification on-ground testing method to accommodate the maintenance programs of enormous and complex ORUs.

This by-step classification method is based on detailed analysis of the tasks, while classified by verified objects, verified items, and verified methods. It can optimize the accuracy, cost, and time consumption of on-ground tests by extended objects and items while choosing the most proper method mentioned in Chapter 3 step by step.

The by-step classification on-ground testing method includes four steps, as follows.

The first step is to focus on typical ORUs, which should have earlier phase verification. These typical ORUs are defined by curbing ORU maintenance type, interface, and the difficulty of identification, usually including energy equipment, information equipment, and fluid pipelines. Visibility and attainability can be verified in this step, which usually takes place throughout computer picture simulations [15].

The second step is to focus on ORUs with difficult operations, such as those with complex operations and narrow operating space, as well as large-scale equipment. Maintenance operations and the scope of impact can be verified in this step for ORU and its tools, which usually occurs throughout computer picture simulation plus on-ground testing (e.g. 1-g simulation, hanging maneuver), as shown in Figure 8.

The third step is to focus on all kinds of ORUs at least with complex operations and small operating space. Maintenance verifications are performed in the simulated cabin environment. The maintenance procedure, operation, time consumption, scope of impact, and certain part supporting can be verified in this step, which usually takes place throughout 1-g simulation, hanging maneuver, and VR.

The fourth step is to face as many ORUs of the space station as possible. All the verified items mentioned in Section 2.2 should be verified. The hanging maneuver, neutral buoyancy tank, VR, and other methods can be used.

Verified items can be classified into operation tests and impact tests. Then, by classifying the testing environment, testing plans and methods can be nailed down (as mentioned by Table 1). The detailed classification method is shown as Table 2.

The verification of maintenance operation is the critical part of maintenance verification. It can mainly prove the ORU maintenance process, operating design, space, tools, and time consumption.

Based on the factors illustrated in Table 1, it is understood that the hanging maneuver and neutral buoyancy tank can simulate 3-D activities in the weightless environment with high accuracy but at a higher cost, which means it cannot perfectly meet the verification needs of the China space station. This paper proposes a simplified zero-gravity simulating partial testing platform that considers the adoptability, effectiveness, and cost of the lifecycle. It can verify the maintenance operation, space, and tools for the internal and external cabin ORU.

The zero-gravity simulating partial testing platform simplifies the internal and external cabin environment. The platform includes the internal cabin environment stimulated platform, external cabin environment stimulated platform, and zero-gravity hanging system for astronauts, products, and tools, as shown in Figures 9-11.

Figure 9 describes the zero-gravity hanging system. The system is designed as a passive balanced system, including a fixed pulley, tether, and balance weight. It has four degrees of freedom, which are upper and lower, front and rear, left and right, as well as rotational freedom.

The internal cabin environment stimulated platform needs to build a cabinet on the ground as shown in Figure 10. The external cabin environment stimulated platform is built like a semicircle arcs cabin, along with obligated common interfaces on the surface as shown in Figure 11.

Different methods of maintenance verification have been used in the maintenance program for the China space station.

As expressed in Table 1, earlier phase verification should be arranged for typical ORUs of different kinds by computer picture simulation. Figure 12 shows the maintenance simulation of a lithium-ion battery. According to Figure 12, it can be concluded that the non-positive operation of ORUs has poor accessibility. Therefore, ORUs with non-positive operation should be verified one by one.

The zero-gravity simulating partial testing platform can be used for the second step verification in Table 2. In fact, this platform has already been practiced in real work. Figure 13 shows the maintenance verification by an engineer.

Then, maintenance verifications in the simulated cabin environment can be arranged. It is not necessary to equip all real products (the same configuration as launched) at the beginning. Figure 14 shows the first round of maintenance operation verification in the internal cabin environment.

After several rounds of improvement, the products with the same configuration launched can be equipped for testing. Then, the fourth step in Table 2 can be arranged to verify all the items using different tests and methods.

The work described in this paper was supported by a grant from the Shanghai Science and Technology Commission Excellent Technical Leader Program Project (No. 17XD1423500).